Copyright TechTarget - All rights reserved ComputerWeekly’s best articles of the day https://cyber.law.harvard.edu/rss/rss.html Techtarget Feed Generator en Sat, 04 Jul 2026 09:29:09 GMT https://www.computerweekly.com editor@computerweekly.com <p>In a boost to the Liberty Global, Telefónica and InfraVia joint venture’s plans to compete with BT Openreach on a more robust basis in the UK’s full-fibre arena, the Competition and Markets Authority (CMA) is to fast-track to Phase 2 of the plans by broadband firm Nexfibre to acquire Substantial Topco and its subsidiary undertakings, including independent broadband providers (altnets) Brsk, YouFibre and, most significantly, <a href="https://www.netomnia.com/">Netomnia</a>.</p> <p>The acquisition – for which the Netomnia element alone is valued at <a title="https://url.us.m.mimecastprotect.com/s/4tdXCyP6mEtJGzyRZcZfwfxP1-0?domain=nexfibre.co.uk/" href="https://url.us.m.mimecastprotect.com/s/4tdXCyP6mEtJGzyRZcZfwfxP1-0?domain=nexfibre.co.uk/">£2bn</a> – was <a href="https://www.computerweekly.com/news/366639123/VMO2-owners-make-Substantial-acquisition-to-gain-altnet-broadband-business">first announced in February 2026</a>, in a deal that Nexfibre claimed would unlock a value of £3.5bn, creating sustainable, scaled network competition and enhanced wholesale choice in the UK.</p> <p>Founded in 2019 and owned by investors Advencap, DigitalBridge and Soho Square Capital, <a href="https://www.substantial.group/">Substantial Group</a>&nbsp;is expected to have more than 3.4 million fibre premises and over 500,000 customers by completion of the deal. Substantial Group’s fibre network, Netomnia, is regarded as the UK’s fourth-largest full-fibre network and second-largest UK altnet.</p> <p>Netomnia&nbsp;undertook a&nbsp;<a href="https://www.computerweekly.com/news/366588862/Altnets-Netomnia-brsk-announce-merger">merger with fellow altnets YouFibre and Brsk in the summer of 2024</a>, and the company ended 2024 with 2.08 million premises serviceable, adding 1.27 million in the year in which it made an acquisition that added 255,000 in the final quarter. It had 238,000 premises connected in the 12-month period, representing 171,000 in a year and 48,000 in Q4 2024. Together with YouFibre and Brsk, Netomnia has had a target to reach five million UK premises serviceable by the end of 2027.</p> <p>Even though the footprint of the UK’s altnets has grown strongly over the past two years, analysts have warned for some time that&nbsp;<a href="https://www.computerweekly.com/news/366636114/UK-altnet-market-entering-its-most-dangerous-phase-yet">companies in the sector were facing increased commercial pressure</a>&nbsp;that could result in a consolidation wave.</p> <p>Commenting on the deal in February 2026, Substantial Group CEO Jeremy Chelot said the “landmark” transaction with Nexfibre – in which <a href="https://www.computerweekly.com/news/366599133/Nexfibre-nears-13-million-premises-passed-reveals-future-build-locations">Virgin Media O2 has a major holding</a> – represented a natural evolution of the UK’s fibre market. “Consolidation has been inevitable, and this deal creates the scaled, sustainable platform needed to drive genuine wholesale competition,” he said.</p> <p>“Importantly, our retail brand, YouFibre, will remain post-close, ensuring our customers continue to receive the same trusted service they know today, while benefiting from the financial strength and infrastructure scale this combination delivers. This is about building a stronger future for UK fibre.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about UK broadband</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366644766/VodafoneThree-extends-CityFibre-partnership-to-support-broadband-services">VodafoneThree extends CityFibre partnership to support broadband services:</a> Independent broadband provider secures new mobile network transmission contract to reach 5G Standalone ambitions&nbsp;supporting retail services, enabling greater capacity, faster speeds and new retail products.</li> <li><a href="https://www.computerweekly.com/news/366644931/ITS-Nufibre-extend-UK-broadband-offerings">ITS, Nufibre extend UK broadband offerings:</a> UK business connectivity full-fibre operator strengthens access to purpose-built business connectivity across England, Scotland and Wales, while fellow altnet looks to make broadband management simpler and more accessible.</li> <li><a href="https://www.computerweekly.com/news/366639950/Altnets-force-to-be-reckoned-with-in-UK-broadband">Altnets ‘force to be reckoned with’ in UK broadband</a>: Research shows peers reaching around 19.7 million premises, with more than 3.5 million live connections, outperforming the major providers on customer satisfaction and value.</li> <li><a href="https://www.computerweekly.com/news/366636114/UK-altnet-market-entering-its-most-dangerous-phase-yet">UK altnet market ‘entering its most dangerous phase yet’</a>: Research predicts shakeout for UK’s independent broadband sector, with likely survivors being those companies that build infrastructure where it matters, invest where returns can be realised and collaborate where possible.</li> </ul> </div> </div> <p>On 11 June, Nexfibre and Substantial <a href="https://www.gov.uk/cma-cases/nexfibre-slash-substantial-merger-inquiry#administrative-timetable">requested the CMA to</a> make a fast-track reference for an in-depth investigation of the takeover at Phase 2.</p> <p>After studying their proposal, the CMA concluded that the conditions to accept a fast-track reference request under section 34ZF(3) of UK competition law have been met, and that it would be appropriate to accept the fast-track reference request and proceed to a Phase 2 investigation. The statutory deadline for Phase 2 of the investigation is 15 December.</p> <p>Offering rationale for hastening the procedure, Nexfibre said its acquisition of Netomnia would be a landmark deal for the UK fibre market, moreover underpinning the country’s ambitions for the future of its digital economy. It added that it was important that the Competition and Markets Authority (CMA) had the analysis and information it needs to reach a timely decision.</p> <p>Furthermore, the provider warned that the current UK altnet market remains fragile, and that the window to secure long-term competition was narrowing. It argued that approving the acquisition would establish wholesale competition to BT Openreach, driving greater choice, better quality and improved value for customers, while supporting a faster transition to full-fibre. The choice, it stressed, was clear: establish a strong competitive framework for next-generation broadband, or maintain what it called the monopolistic status quo holding the country back.</p> <p>“We requested a fast-track to Phase 2 to get to the right answer faster, ensuring due process while recognising urgency,” said Nexfibre CEO Rajiv Datta. “We look forward to continuing our constructive engagement with the CMA.</p> <p>“This deal would create the scaled, sustainable alternative to the BT Openreach monopoly, something the UK market still lacks. Every day of delay reinforces the incumbent’s advantage and slows the progress of genuine competition.”</p> <p>Yet Nexfibre’s rivals have warned that the takeover raises substantial questions about the competitiveness of the UK’s communications market.</p> <p>Commenting on the CMA’s move, Simon Holden, <a href="https://www.computerweekly.com/news/366643869/CityFibre-breaks-one-million-connections-barrier">CityFibre</a> CEO, said: “VMO2/Nexfibre’s planned acquisition of Netomnia would remove a successful challenger and reduce choice for consumers. With 80% overlap between the two networks, the deal raises significant questions, and the CMA is right to take an in-depth look at its impact on UK digital infrastructure and the competition that policymakers, regulators and the altnets are working so hard to establish.”</p> UK regulatory watchdog approves next phase of UK broadband provider’s acquisition of leading independent company, but rivals warn of diminished competition https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/Netomnia-hero.jpg https://www.computerweekly.com/news/366645270/UK-competition-authority-fast-tracks-Nexfibre-altnet-acquisition Thu, 02 Jul 2026 10:30:00 GMT <p>On 12 June 2026, the US Department of Commerce issued an emergency directive that forced Anthropic <a href="https://www.computerweekly.com/news/366644826/US-suspension-of-Anthropic-models-prompts-AI-sovereignty-calls">to suspend foreign nationals</a> from using Claude Fable 5. It came suddenly with no exemption for allied nations and no appeals process. One jurisdiction flipped a switch, and the world’s access to a leading piece of frontier artificial intelligence (AI) went dark.</p> <p>The ban, which the US has only <a href="https://www.bbc.co.uk/news/articles/cdr42623e1do">now begun to lift</a> after weeks of diplomatic fallout, marks more than an AI safety controversy. It crystallises a question that European and UK policymakers have been racing to answer through an unprecedented wave of legislative and industrial interventions. Namely, when a single nation controls the infrastructure upon which your economy, public services and national security depend, what does sovereignty actually mean – and can you build it quickly enough?</p> <p>This article examines whether Europe and the UK can <a href="https://www.computerweekly.com/feature/Is-cloud-data-sovereignty-all-just-a-case-of-Trust-me-bro">close a digital sovereignty gap</a> that decades of dependence on American technology have rooted in the continent’s economic foundations. Drawing on economic data, historical precedent and academic analysis, it looks at whether catch-up remains possible and tests that against three prerequisites demonstrated by successful late-developer states in modern history. These are unified strategic funding at a scale that matches the competition, the willingness to use internal markets as an exclusionary weapon, and the humility to copy rather than insist on innovating.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-techGDP-800px-f.jpg"> <img data-src="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-techGDP-800px-f_mobile.jpg" class="lazy" data-srcset="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-techGDP-800px-f_mobile.jpg 960w,https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-techGDP-800px-f.jpg 1280w" height="436" width="559"> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <section class="section main-article-chapter" data-menu-title="Urgency in Brussels and Westminster"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Urgency in Brussels and Westminster</h2> <p>The recent flurry of policy announcements suggests that Brussels and Westminster are gripped by some urgency.&nbsp;</p> <p>The European Union’s (EU) <a href="https://www.computerweekly.com/news/366643862/EU-unveils-full-stack-sovereignty-package-to-build-Euro-tech-muscle">Cloud and AI Development Act</a> – part of the European Commission’s AI Continent Action Plan – proposes a single EU-wide sovereignty framework with four escalating levels of autonomy, from basic physical data residency (Level 1) through to full software supply chain transparency and demonstrated independence from third-country interference (Level 4). The act also targets a threefold <a href="https://www.computerweekly.com/news/366638951/Power-supply-issues-flagged-as-major-growth-inhibitor-of-European-datacentre-market">expansion of European datacentre capacity</a> by the 2030s, backed by streamlined permitting and improved access to energy, land and financing.</p> <p>Alongside this sits the Chips Act 2.0, which shifts the EU’s semiconductor strategy from supply-focused investment to demand cultivation. Its predecessor mobilised more than €52bn in public and private investment and created an estimated 46,000 jobs. The sequel introduces “grand challenges” aimed at industrial development of AI-critical chips, demand accelerators to speed products to market, and state aid provisions for first-of-a-kind fabrication projects not yet present in the Union.</p> <p>The <a href="https://www.computerweekly.com/news/366643862/EU-unveils-full-stack-sovereignty-package-to-build-Euro-tech-muscle">EU Open Source Strategy</a> completes the triad, mandating public administrations to act as anchor users of open source alternatives – cloud, workplace tools, secure email – and developing a sovereign “Open Internet Stack” designed to displace proprietary foreign software from government workflows.</p> <p>The UK, operating outside EU mechanisms but facing identical structural vulnerabilities, has assembled its own arsenal. The £1.1bn <a href="https://www.computerweekly.com/news/366643844/Starmer-announces-sovereign-compute-strategy-amid-11bn-chip-investment">sovereign compute strategy</a> – announced by technology minister Liz Kendall with the declaration that “AI is the defining currency of economic and hard power in today’s world” – allocates £750m to a national AI supercomputer, including £400m for specialist chip procurement. A £120m AI hardware innovation programme funds domestic chip design and testing. The <a href="https://www.computerweekly.com/news/366641682/UK-governments-50m-sovereign-AI-fund-bids-to-commercialise-research">£500m Sovereign AI Unit</a>, launched in April 2026, aims to become the primary investment hub for high-potential UK AI startups, while a separate £500m <a href="https://www.computerweekly.com/news/366642362/Government-funds-self-learning-AI-company">sovereign AI fund targets growth-stage companies</a> with investments between £1m and £10m.</p> </section> <section class="section main-article-chapter" data-menu-title="The key levers of digital sovereignty"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The key levers of digital sovereignty</h2> <p>These measures can be grouped into four strategic levers that the EU and UK are pulling simultaneously:</p> <ol class="default-list"> <li><strong>Supply-side industrial policy</strong> – the state directly funding physical infrastructure to bypass market reliance on foreign hyperscalers.&nbsp;</li> <li><strong>Procurement-led sovereignty</strong> – using public sector purchasing power to set market standards and force technical requirements on suppliers, from open source mandates to local data residency.&nbsp;</li> <li><strong>Regulatory shielding</strong> – establishing legal frameworks, such as France’s SecNumCloud 3.2 qualification and Germany’s BSI C3A two-tier cloud autonomy criteria, that force foreign providers to meet sovereignty definitions or face exclusion from government workloads.&nbsp;</li> <li><strong>Innovation ecosystem support</strong> – targeted grants and investment funds designed to cultivate homegrown intellectual property.</li> </ol> <p>The cumulative investment is substantial on paper. But it must be measured against a global technology economy in which the imbalance of power is not only large but structural.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-top20cos-800px-f.jpg"> <img data-src="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-top20cos-800px-f_mobile.jpg" class="lazy" data-srcset="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-top20cos-800px-f_mobile.jpg 960w,https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-top20cos-800px-f.jpg 1280w" height="421" width="560"> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="Data shows US digital dominance"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Data shows US digital dominance</h2> <p>Computer Weekly analysis of audited annual reports and regulatory filings, benchmarked against World Bank and IMF GDP statistics, reveals that US-headquartered technology firms generated combined global revenues exceeding $2tn in their most recent fiscal years.&nbsp;</p> <p>Microsoft alone posted $281.7bn – more than the combined tech revenue of every major firm headquartered in France, Italy, Spain and the UK, and rivalling Germany’s entire listed tech sector.&nbsp;</p> <p>Amazon Web Services (AWS) contributed $128.7bn in cloud infrastructure revenue, Alphabet $402.8bn, and Nvidia – the semiconductor designer whose graphics processing units (GPUs) power the AI revolution – $130.5bn.</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-scatterEUUK-800px-f.jpg"> <img data-src="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-scatterEUUK-800px-f_mobile.jpg" class="lazy" data-srcset="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-scatterEUUK-800px-f_mobile.jpg 960w,https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-scatterEUUK-800px-f.jpg 1280w" height="363" width="558"> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>By contrast, Europe’s largest technology companies operate at a fraction of that scale. Germany’s SAP, the continent’s flagship enterprise software firm, reported $41.6bn. France’s Dassault Systèmes reached $7.4bn. The UK’s ARM – a globally significant semiconductor intellectual property (IP) designer – generated $4bn.&nbsp;</p> <p>No European-headquartered company operates a hyperscale cloud infrastructure business at a level competitive with AWS, Microsoft Azure or Google Cloud. The delivery model that underpins the modern digital economy – elastic, on-demand compute – is almost entirely American-owned.</p> <p>Tech GDP as a percentage of national output tells the same story from a different angle.&nbsp;</p> <p>The US tech sector accounts for 7.1% of GDP against a $30.5tn economy. The UK sits at 5%, Germany at 5.1%, France at 5.2%. The gap appears modest until measured against the countries that have already caught up. South Korea registers 12.3% tech GDP – the highest of any advanced economy – built on Samsung Electronics ($233.3bn), SK Hynix ($68.7bn), and a state apparatus that for decades directed credit to specific firms through state-directed credit mechanisms. Taiwan hits 12% on the back of TSMC ($122.4bn) and Foxconn ($213.7bn), companies incubated under conditions of strategic state protection and explicit government coordination.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Data sources and methodology</h3> <p>The selection of companies for this analysis was designed to map the foundational architecture of the global digital economy, and gave us a “global tech power list” of 80 firms.&nbsp;</p> <p>For US firms, we identified the top 20 by annual revenue, which established a baseline threshold of approximately $20bn.&nbsp;</p> <p>For international markets, we applied this same $20bn revenue floor as a criterion where possible. But, because the prevalence of enterprise-scale technology firms varies significantly by region, we had to transition to selecting the leading firms in each country that operate within the tech ecosystem.&nbsp;</p> <p>While our primary focus remained on foundational ICT suppliers – ie, companies that provide the cloud, software, semiconductor and enterprise infrastructure on which the digital economy relies – we have included major electronics manufacturers and telecommunications providers. This reflects that in modern tech supply chains, the boundary between pure-play digital and physical manufacturing and connectivity layers has become porous.</p> <p>Data on corporate revenue is derived from audited annual reports and regulatory filings. National GDP statistics are sourced from international benchmarks provided by the World Bank and IMF. Tech GDP estimates represent combined economic value added by the ICT sector, calculated based on sources that include the OECD, Eurostat, national government reports and the US International Trade Administration.</p> </div> </div> </section> <section class="section main-article-chapter" data-menu-title="The state as driver of development"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The state as driver of development</h2> <p>What distinguishes the South Korean and Taiwanese outcomes from European aspirations is not merely the scale of investment but the political architecture that directed it.&nbsp;</p> <p>Andrew Wright, a lecturer at Hult International Business School who specialises in international political economy, frames the distinction through the lens of historical catch-up dynamics.&nbsp;</p> <p>“Germany and America overtook British leadership through protectionism and state promotion of industry,” notes Wright. “As a result, they developed giant monopolistic – or, more accurately, oligopolistic – corporate giants. Scale was important then and still is.”</p> <p>The post-war generation of East Asian developers – Japan, then Korea, then Taiwan – intensified this model.&nbsp;</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.computerweekly.com/rms/computerweekly/CW-analysis-Antony-Adshead-HeatmapRevenueByRegionAndSector-800px-f.jpg"> <img data-src="https://www.computerweekly.com/rms/computerweekly/CW-analysis-Antony-Adshead-HeatmapRevenueByRegionAndSector-800px-f_mobile.jpg" class="lazy" data-srcset="https://www.computerweekly.com/rms/computerweekly/CW-analysis-Antony-Adshead-HeatmapRevenueByRegionAndSector-800px-f_mobile.jpg 960w,https://www.computerweekly.com/rms/computerweekly/CW-analysis-Antony-Adshead-HeatmapRevenueByRegionAndSector-800px-f.jpg 1280w" height="270" width="559"> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> <p>“These countries used much more extensive state support, protection, and, frankly, overall direction,” Wright explains. “All of these states borrowed and stole technology, promoted giant firms, protected home markets and largely copied advanced nations while benefiting from lower costs.”&nbsp;</p> <p>The pattern is so consistent that the economic historian Alexander Gerschenkron identified a typology: the later the catch-up project, the greater the role played by the state. China, the most recent entrant, represents the apotheosis of this trajectory – state capitalism at continental scale, with Alibaba ($137.3bn), Tencent ($104.5bn) and Huawei ($126bn) now rivalling Western incumbents on revenue and increasingly on technological capability.</p> <p>But the East Asian precedent carries a complicating element for Europe.&nbsp;</p> <p>Cheap labour – the mechanism that enabled Korea and Taiwan to drive export-led growth while protecting domestic markets – is structurally unavailable to advanced European economies.&nbsp;</p> <p>“Europe cannot emulate this in competing with the US,” Wright argues. The more relevant parallel lies further back in history. “In the late 19th century, low wages were not a major factor in the rise of the US and Germany. It was more about state tutelage, the growth of the giant corporation, and the systematisation of research and development.”</p> <figure class="main-article-image full-col" data-img-fullsize="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-allCos-800px-f.jpg"> <img data-src="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-allCos-800px-f_mobile.jpg" class="lazy" data-srcset="https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-allCos-800px-f_mobile.jpg 960w,https://www.computerweekly.com/rms/computerweekly/CW-analysys-Antony-Adshead-allCos-800px-f.jpg 1280w" height="902" width="560"> <div class="main-article-image-enlarge"> <i class="icon" data-icon="w"></i> </div> </figure> </section> <section class="section main-article-chapter" data-menu-title="Responding at scale hampered by fragmentation"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Responding at scale hampered by fragmentation&nbsp;</h2> <p>This distinction proves critical when evaluating the three prerequisites against Europe’s present position.</p> <p>The first prerequisite – unified strategic funding at a scale that matches American finance – faces the immediate obstacle of fragmentation. The UK’s £1.1bn sovereign compute programme and the EU’s €52bn Chips Act mobilisation represent genuine political commitment, but they operate through entirely separate governance structures, procurement pipelines and strategic priorities.&nbsp;</p> <p>South Korea’s window guidance system, by which the central bank literally directed private banks to extend credit to designated firms, required a concentration of financial authority that no European institution – not the European Investment Bank, not the European Central Bank, not any national treasury – currently possesses.&nbsp;</p> <p>“Only a hugely concentrated strategic mechanism of funding can compete with what the Americans now have available to them,” Wright observes, “partly due to their past and partly due to the sheer size of American finance.” He adds: “Current efforts to seed little developmental pockets, or the shoots of venture capitalism, cannot possibly compete.”</p> </section> <section class="section main-article-chapter" data-menu-title="Using the market: The Airbus example"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Using the market: The Airbus example</h2> <p>The second prerequisite – the willingness to use Europe’s internal market as an exclusionary weapon – has a proven precedent.&nbsp;</p> <p>Airbus, the European aerospace consortium, now competes directly with Boeing after decades of state-backed development that American free-market economists once derided as incapable of picking winners.&nbsp;</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> I would see Airbus as a model. If I was a pro-EU policy maker, I would stop American private companies from feeding off funds and contracts of the European states, and replace those relations with European contractors </figure> <figcaption> <strong>Andrew Wright, Hult International Business School</strong> </figcaption> <i class="icon" data-icon="z"></i> </div> </blockquote> <p>“I would see Airbus as a model,” says Wright. The logic transfers directly to digital infrastructure: “If I was a pro-EU policy maker, I would stop American private companies from feeding off funds and contracts of the European states, and replace those relations with European contractors.”</p> <p>Elements of this approach are already visible. The UK’s Science, Innovation and Technology Committee has proposed a “cloud consumption dashboard” to publicly track contract awards by supplier, mandating SME spending targets and break clauses in contracts with foreign providers. The Procurement Act 2023 is being updated to require public sector bodies to prioritise open source solutions. The EU’s four-level sovereignty framework creates a ratcheting mechanism by which providers must demonstrate progressively deeper independence from third-country jurisdiction to qualify for sensitive workloads.</p> <p>The counter-current, however, is powerful. Decades of procurement decisions have entrenched American hyperscalers within European public sector architecture.&nbsp;</p> <p>The network effects are self-reinforcing. The more departments standardise on a single cloud platform, the harder and more expensive exiting becomes.&nbsp;</p> <p>Wright describes this as “path dependence” – the accumulated weight of past choices that makes deviation costly even when the strategic case for change is overwhelming. “Europe has existed under American tutelage, allowing US firms access and the development of dominance,” he notes. “Breaking out of this is costly and difficult.”</p> </section> <section class="section main-article-chapter" data-menu-title="Copy, don’t innovate"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Copy, don’t innovate</h2> <p>The third prerequisite – the willingness to copy rather than insist on innovating – cuts against Europe’s self-conception as a centre of original research.&nbsp;</p> <p>Wright points to a historical irony. “Much of the technology that has fed the American machine was originally developed in Europe. Yet Europe has often been unable to control and commodify it, or build the giant dominant firms that emerged from it,” he says.&nbsp;</p> <p>The East Asian developers faced no such cultural constraint. “They rarely innovated; they mostly copied,” says Wright.</p> <p>The EU Open Source Strategy represents an implicit acknowledgement of this logic – building a catalogue of open alternatives that, by definition, replicate proprietary functionality rather than inventing from scratch. The Gaia-X federated infrastructure initiative creates interoperability standards designed to let local European providers link services in direct competition with non-EU hyperscalers, again a copying-and-competing model rather than a leapfrogging one.&nbsp;</p> <p>Whether European political culture can sustain this pragmatic posture – building what works rather than what is original – remains an open question.</p> </section> <section class="section main-article-chapter" data-menu-title="Digital sovereignty at a crossroads"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Digital sovereignty at a crossroads</h2> <p>The question that faces European policymakers is not whether digital sovereignty is desirable but whether the window for achieving it remains open.&nbsp;</p> <p>The Fable 5 episode demonstrates that dependency is not a theoretical risk that can be managed through contractual fine print. It is an operational reality that can be triggered by a foreign government’s unilateral decision, with consequences that cascade through every service, every department, and every citizen who relies on infrastructure they do not control.</p> <p>The revenue data suggests the structural gap is widening, not closing. American technology firms grow at rates European competitors cannot match, fuelled by a venture capital ecosystem, a stock market depth and a military-industrial procurement pipeline that Europe has no equivalent of.&nbsp;</p> <p>The US, Wright argues, operates a “hidden developmental state” – less visibly directive than the East Asian model but no less effective, leveraging defence spending, public sector contracts and deep state-corporate integration to sustain technological dominance while maintaining the rhetorical posture of free enterprise.</p> <p>The conditional verdict, then, is that Europe and the UK can catch up – but the path demands a level of political coordination, financial concentration and strategic pragmatism that the continent has rarely mustered outside of wartime mobilisation or the Airbus project.&nbsp;</p> <p>The measures now being deployed are genuine in intent and substantial in scale by historical European standards. Whether they are sufficient in speed and coordination to overcome decades of accumulated dependency is the question on which the next decade of European digital sovereignty will turn.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about data sovereignty and cloud infrastructure</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/feature/Is-cloud-data-sovereignty-all-just-a-case-of-Trust-me-bro">Is cloud data sovereignty all just a case of ‘Trust me, bro’?</a> An analysis of whether contractual sovereignty guarantees from hyperscale cloud providers offer meaningful protection or merely cosmetic reassurance to European customers.</li> <li><a href="https://www.computerweekly.com/feature/This-rise-of-the-splinternet-Data-sovereignty-risks-and-responses">The rise of the splinternet? Data sovereignty risks and responses</a>; We look at the political, legal, and economic risks that data sovereignty fractures pose to global digital infrastructure, from sanctions-driven cloud lockouts to jurisdictional conflicts.</li> </ul> </div> </div> </section> As the US demonstrates it can wield an AI ‘kill switch’, the EU and UK unleash a wave of sovereign tech measures. Can state-led industrial policy bridge a $2tn revenue chasm? https://cdn.ttgtmedia.com/visuals/German/article/acquisition-takeover-adobe.jpg https://www.computerweekly.com/news/366645336/Data-dive-Kill-switch-and-catch-up-can-Europe-close-the-sovereignty-gap Thu, 02 Jul 2026 06:15:00 GMT <p>As part of assisting the German emergency service’s clear stated mission to extinguish fires before they start, network cabling provider from <a href="https://www.tde.de/index-en">Trans Data Elektronik</a> (TDE) has revealed it has been working with the <a href="https://www.berliner-feuerwehr.de/fileadmin/bfw/dokumente/Strategie-2030/20251211_BFw_Strategy-2030_EN_Screen.pdf">Berlin Fire and Rescue Service</a> to upgrade the core network on which its operations are based, guaranteeing security and reliability at every level and elevating capability in the server room.</p> <p>Responding to more than 500,000 calls a year and rising, the Berlin Fire and Rescue Service plays a key role in keeping the German capital running safely day in, day out. It employs more than 6,600 people in fire-fighting and rescue service, administration and trainee programmes to handle the immense workload, operating out of 35 fire stations spread across the city.</p> <p>In addition to handling fire emergencies, emergency medical services make up the vast majority of its operations at 86%, followed by technical assistance and reconnaissance.</p> <p>To guarantee smooth operations, the service requires a high-performance and stable IT infrastructure to reliably respond to emergency calls and coordinate operations. However, over time, the department has accumulated a multitude of different systems and add-ons in its server racks – a complex environment that is regarded as no longer meeting current requirements.</p> <p>Indeed, the server cabinets dated from 1991 and 1992, and due to numerous retrofitted cables and components, could only be opened with great difficulty and by a dedicated team of technicians.</p> <p>There was no active ventilation; the air conditioning was too noisy and incorrectly sized; and the power distribution system was limited to a simple power strip. Copper, fibre-optic patch and power cables would lie haphazardly side by side on the cable rack, and a lack of cable management systems resulted in a confusing tangle of cables. This made maintenance work, upgrades and repairs considerably more difficult.</p> <p>Essentially, the condition of the server room did not meet the required security standards in any respect – neither operationally or physically nor in terms of information technology. A comprehensive overhaul was essential.</p> <section class="section main-article-chapter" data-menu-title="Operational readiness"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Operational readiness</h2> <p>As the Berlin Fire and Rescue Service is classified as a public authority or organisation with security responsibilities, the IT infrastructure could not simply be shut down and replaced. The system had to remain operational while the fire department continued its work to ensure its unrestricted operational readiness.</p> <p>To implement the upgrade project, the Berlin Fire and Rescue Service decided to bring in professional assistance from <a href="https://www.tde.de/index-en">Dortmund-based</a> TDE, which was responsible for supplying the <a href="https://www.computerweekly.com/news/366632286/Nokia-launches-FTTH-digital-twin-AI-tools-to-boost-network-reliability">passive network</a> components. The team also supported the fire and rescue service in selecting the system, planned the configuration of the racks and commissioned the company <a href="https://www.wisag.de/en/">Wisag</a> to ensure the control room was equipped in accordance with health and safety regulations.</p> <p>The passive network infrastructure of the server room was completely renewed and consolidated to create a standards-compliant, structured and highly available foundation for IT operations. TDE rolled its TML system – a modular cabling system featuring pull-out 19-inch module carriers, each holding eight modules, micro-distribution internal cables with 12F-MPO, TP trunk cables and MPO/MTP connections.</p> <p>The system is designed to allow TP and fibre-optic modules to be combined in a single module carrier, thereby ensuring what TDE claims is “exceptionally high” packing density of up to 96 fibre-optic ports or 48 copper ports in just one height unit with the MPO system.</p> <p>The cabling using TML micro-distribution indoor cables and eight 12-fibre MPO connectors in particular, each serving six ports, is said to offer a significant advantage over conventional services using individual patch cables, breakout cables or in-house cabling. To ensure reliability and avoid disrupting day-to-day operations, the fire service set up the passive network infrastructure in parallel with the existing system and then migrated the services in stages.</p> <p>The cabling firm accepts that meeting the fire and rescue service’s high standards was a challenge. In addition to certified quality, reliable partners, short response times and personalised advice, the costs also played an important role.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about emergency services communications</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366645034/HMD-Secure-Qualcomm-StreamWIDE-forge-NR-Sidelink-collaboration">HMD Secure, Qualcomm and StreamWIDE&nbsp;forge&nbsp;NR Sidelink collaboration</a>: Partnership unveils NR Sidelink-powered device-to-device communications boasting tech firsts for first responders and critical infrastructure teams operating beyond network coverage<a href="https://www.computerweekly.com/news/366645034/HMD-Secure-Qualcomm-StreamWIDE-forge-NR-Sidelink-collaboration">Read More</a></li> <li><a href="https://www.computerweekly.com/news/366645163/Contrivian-broadens-Horizon-for-first-response-connectivity">Contrivian broadens Horizon for first response connectivity</a>: Mission-critical enterprise and government comms provider claims to be redefining connectivity, bringing reliable field communications with&nbsp;intelligent&nbsp;5G/LTE&nbsp;routing and SD-WAN.</li> <li><a href="https://www.computerweekly.com/news/366640616/Avaya-finds-louder-voice-for-mission-critical-comms-platform">Avaya finds louder voice for mission-critical comms platform</a>: Enterprise software provider&nbsp;unveils mission-critical platform&nbsp;to deliver secure, zero-downtime voice communications for environments including government, healthcare, financial and emergency services.</li> <li><a href="https://www.computerweekly.com/news/366644772/Motorola-Solutions-looks-to-boost-public-safety-with-expanded-Assist-AI">Motorola Solutions looks to boost public safety with expanded Assist AI</a>: Tech firm rolls out mission-critical AI for public safety use cases to reduce administrative burden and accelerate incident response.</li> </ul> </div> </div> <p>Following four years of planning, refurbishment, fit-out and installation, the server room went live with its new equipment at the end of 2025, and is said to have stabilised the technical infrastructure of the IT system.</p> <p>The passive infrastructure has now been consolidated and brought up to the latest standards, and TDE assured that it now forms a stable, standards-compliant and reliable foundation for IT operations. In particular, availability and fault tolerance are said to have been noticeably improved.</p> <p>In all, TDE said the modern rack system, the structured cabling and the clear separation of the different systems can minimise the risk of failure and unplanned interruptions.</p> <p>Furthermore, it said the new structure increases transparency and enables clear documentation, allowing for analysing faults more quickly and maintenance or expansion work to be carried out without disruption while the system is in operation.</p> <p>The homogeneous network expansion also reduces the training requirements on staff and allows for more efficiently and streamlined spare parts management. Ultimately, the modernisation has laid the groundwork for future scaling, achieving higher power densities and retrofitting technological advancements in the datacentre environment.</p> <p>“We have been working with TDE on passive network infrastructure since 2021,” said Marco Weise, ICT operations manager at the Berlin Fire and Rescue Service.</p> <p>“That’s why it was important to us that the new system was fully compatible with our existing solution. In this project, TDE took on the primary responsibility for planning, supply and order processing of the entire passive infrastructure and fulfilled this task to our complete satisfaction. We are very pleased with the modernisation. We will continue to rely on TDE’s expert advice for future projects.”</p> </section> Looking to ensure that rescue and fire-fighting operations continue to run seamlessly in the future, Germany’s largest professional fire department modernises passive network infrastructure https://cdn.ttgtmedia.com/visuals/German/article/berlin-1-adobe.jpg https://www.computerweekly.com/news/366645302/Berlin-Fire-and-Rescue-Service-enhances-passive-network-infrastructure Thu, 02 Jul 2026 05:00:00 GMT <p>The emerging artificial intelligence (AI) hubs of Mumbai and Chennai in India and Singapore are said to be among Asia’s leading cloud and AI ecosystems, and to tap into this highly dynamic market and strengthen its connectivity offerings, global communications technology player <a href="http://www.tatacommunications.com/">Tata Communications</a> has made strategic investments in subsea cable infrastructure in the region via its acquisition of “significant” fibre capacity.</p> <p>The company believes the India-Singapore subsea route is set to become one of the world’s most critical digital corridors in future, representing a high-capacity, low-latency pathway that will underpin critical enterprise, cloud and hyperscaler traffic between India, Southeast Asia and global markets.</p> <p>Through these investments Tata Communications says it’s enhancing the Tata Global Network (TGN) as part of a general aim of addressing the growing bandwidth and AI-driven data demands of enterprises across Asia and further extension globally.</p> <p>By enhancing capacity on its TGN network, Tata Communications says it’s furthering its ability to deliver diverse, agile and high-performance connectivity to customers.</p> <p>Specifically, the company will be integrating a subsea cable system between Mumbai and Singapore; investing as a consortium member in a new subsea cable system, connecting Chennai to Singapore with expected ready for service (RFS) in the fourth quarter of 2029.</p> <p>Together, these investments are seen as being able to meet the growing needs of the datacentre ecosystem offering enterprises a scalable, reliable and future-ready connectivity between India and Singapore.</p> <p>The cable systems will further connect with Tata Communications India Terrestrial fibre network for onward connectivity to other parts of the country and to more than 100 datacentres nationwide.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about subsea communications</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366643922/ViaTunisia-subsea-segment-reaches-Ready-for-Service-status">ViaTunisia&nbsp;subsea segment reaches ready-for-service status</a>: Subsea cable segment connecting Marseilles in&nbsp;France and Bizerte&nbsp;in&nbsp;Tunisia has officially reached ready-for-service status, making leap from infrastructure design to live connectivity.</li> <li><a href="https://www.computerweekly.com/news/366640741/Colt-announces-subsea-terrestrial-network-routes">Colt announces subsea, terrestrial network routes</a>: Digital infrastructure company reveals plans to launch international connectivity routes connecting the US West Coast to Asia, marking the latest phase of its major global network expansion.</li> <li><a href="https://www.computerweekly.com/news/366643009/Via-Africa-subsea-cable-project-to-strengthen-European-African-connectivity">Via Africa subsea cable project to strengthen European, African connectivity</a>: Europe-Africa submarine cable project backed by consortium model designed to connect Europe to Africa along the Atlantic coast, enhancing the resilience and diversity of West Africa’s international connectivity.</li> <li><a href="https://www.computerweekly.com/news/366637186/Subsea-cable-worth-1bn-to-link-Japan-with-Malaysia-and-Singapore">Subsea cable worth $1bn to link Japan with Malaysia and Singapore</a>: The Intra-Asia Marine Cable will deliver 320Tbps capacity across the region, complementing subsea cable investments by hyperscalers such as Google and Meta in recent years.</li> </ul> </div> </div> <p>Combined with the global TGN Subsea network, the overall network is seen as enhancing the capabilities of the full suite of <a href="https://www.computerweekly.com/news/366640695/Tata-Communications-unveils-self-healing-network">IZO connectivity offerings</a> such as IZO DC Dynamic Connectivity and IZO multi-cloud providing self-healing, always-on and self-provisioning capabilities across datacentres and cloud ecosystems. Tata notes that customers will be able to activate and integrate these capacities with agility into their networks on demand.</p> <p>“As global demand for digital and AI-driven services continues to accelerate, these investments reinforce our commitment to building future-ready digital infrastructure at scale,” said Genius Wong, Tata Communications’ <a href="https://www.reuters.com/world/india/indias-tata-communications-names-new-chief-technology-officer-2026-06-30/">departing</a> executive vice-president of core and next-gen connectivity services and chief technology officer.</p> <p>“By combining subsea capacity enhancement with both short-term and long-term strategic investments, we are strengthening the reliability, scalability and performance of connectivity solutions for our customers across one of the world’s busiest digital corridors,” she said. “These enhancements align with Tata Communications’ long-term strategy to expand its global subsea network footprint, provide business outcome solutions to customers and reinforce India’s position as a Digital Hub.”</p> <p>The backbone of the portfolio is the Tata Communications Network Fabric subsea fibre network, which comprises more than 500,000km of subsea optical fibre and in excess of 200,000km of terrestrial fibre. To enhance further enterprise connectivity across Asia and beyond, in 2025, Tata Communications has integrated the new Tata Global Network – Intra-Asia 2 (TGN IA2) submarine cable.</p> <p>The latter is said to improve latency for faster performance, enhancing reliability through greater redundancy and increasing network diversity through interconnection with TGN IA.</p> AI-ready connectivity investments look to enable enterprises to connect across continents with secure, high-speed, dynamic, self-managed and low-latency infrastructure https://cdn.ttgtmedia.com/rms/computerweekly/Singapore-1-fotolia.jpg https://www.computerweekly.com/news/366645265/Tata-Communications-strengthens-India-Singapore-connectivity-corridor Wed, 01 Jul 2026 11:30:00 GMT <p>With the UK independent broadband provider sector seeing increased business pressures, two altnets in the country – <a href="http://www.itstechnologygroup.com/">ITS</a> and <a href="https://www.nufibre.co.uk/">Nufibre</a> – have announced steps to improve customer services.</p> <p>As part of the next phase of its investment in its national network built specifically for business use, B2B full-fibre&nbsp;operator ITS has announced it is extending its full-fibre&nbsp;network into 13 additional towns and cities across England, Scotland and Wales.</p> <p>The move is designed to bring more organisations within reach of <a href="https://www.computerweekly.com/news/366642976/Thousands-of-Essex-premises-to-gain-upgraded-broadband">high-capacity connectivity</a>, giving businesses greater choice and creating more opportunity for ITS partners to serve customers that need resilient, scalable services to support cloud, artificial intelligence (AI) and other data-intensive applications.</p> <p>Coventry, Huddersfield and Loughborough will be the first locations to be ready for service&nbsp;later in June 2026. Aberdeen, Bath, Bournemouth, Cardiff, Chester, Edinburgh, Exeter, Plymouth, Southampton and Swansea will go live in phases between July and October.</p> <p>Services on offer range from full-fibre business broadband to dedicated Ethernet. This includes an Ethernet-over-<a href="https://www.computerweekly.com/news/366640022/CityFibre-AllPoints-Fibre-introduce-multi-gigabit-FTTP">fibre to the premises</a> bundled service which combines symmetrical speeds, burstable bandwidth, enhanced support and security at the network edge.</p> <p>“Organisations across the UK need connectivity that is built around how they operate, not a residential model adapted for business use,” said ITS CEO Daren Baythorpe. “That is why we are continuing to invest where the demand is clear and where our digital infrastructure can create long-term value.</p> <p>“This expansion takes us into even more towns and cities across England, Scotland and Wales, supporting the places that <a href="https://www.computerweekly.com/news/366638559/Gigabit-accelerates-across-metro-rural-UK">power local and regional economies</a>, while giving businesses greater choice in areas where high-capacity connectivity is becoming increasingly important,” he added. “It is also about the way the market is served. Partners … need a wholesale platform, a strong product portfolio and the service experience to support customers with confidence, including larger, multi-site organisations that need consistent connectivity across different locations.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about UK altnets</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366639950/Altnets-force-to-be-reckoned-with-in-UK-broadband">Altnets ‘force to be reckoned with’ in UK broadband</a>: Research shows peers reaching around 19.7 million premises, with more than 3.5 million live connections, outperforming the major providers on customer satisfaction and value.</li> <li><a href="https://www.computerweekly.com/news/366636114/UK-altnet-market-entering-its-most-dangerous-phase-yet">UK altnet market ‘entering its most dangerous phase yet’</a>: Research predicts shakeout for UK’s independent broadband sector, with likely survivors being those companies that build infrastructure where it matters, invest where returns can be realised and collaborate where possible.</li> <li><a href="https://www.computerweekly.com/news/366623603/UK-altnets-long-term-survival-in-question-as-96-percent-consider-MA">UK altnets’ long-term survival ‘in question’ as 96% consider M&amp;A</a>: Study from business connectivity firm warns that while alternative broadband providers have fuelled full-fibre expansion and delivered much-needed competition, many are having to explore strategies for survival.</li> <li><a href="https://www.computerweekly.com/news/366639123/VMO2-owners-make-Substantial-acquisition-to-gain-altnet-broadband-business">VMO2 owners make Substantial acquisition to gain altnet broadband business</a>: Acquisition of UK’s second-largest altnet by broadband major claimed to be able to unlock value of £3.5bn, creating sustainable, scaled network competition and wholesale choice in the UK.</li> </ul> </div> </div> <p>The roll-out is also supported by wider investment across ITS’s wholesale platform, and the firm is continuing to develop its self-service portal to give partners greater control and visibility, from order and delivery through to in-life service management.</p> <p>Improved self-service is also at the heart of Nufibre’s official launch of the nuHub v1.0 customer portal, designed and developed entirely in-house to make broadband management simpler and more accessible for customers.</p> <p>Nufibre noted that one of its core goals is to empower customers to manage their broadband services quickly and easily. The nuHub is designed to provide a centralised platform where customers can access and manage key account information, including contact details; broadband packages; billing information; payment methods; router credentials; and IP address information.</p> <p>Features and enhancements will be added to the platform throughout 2026. Already in development for future versions are a fully integrated ticketing system, referral programmes and router self-diagnostics. As part of what it says is a commitment to simplicity, Nufibre has removed the need for traditional passwords, allowing customers to securely access their accounts using a propriety nuPIN verification system.</p> <p>“By giving customers instant access to the information and tools they need, 24/7, we’re enabling them to manage everyday tasks themselves while allowing our support teams to focus on areas where assistance is most needed,” said Nufibre CEO Martin Gardner.</p> UK business connectivity full-fibre operator strengthens access to purpose-built business connectivity across England, Scotland and Wales, while fellow altnet looks to make broadband management simpler and more accessible https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/fibre-broadband-FTTP-adobe.jpeg https://www.computerweekly.com/news/366644931/ITS-Nufibre-extend-UK-broadband-offerings Mon, 29 Jun 2026 04:30:00 GMT <p>Extreme Wireless has revealed that US sports stadia will soon be making use of its next-generation Multi-Beam Wireless stadium connectivity solution which, in combination with MatSing's lens antenna technology, is said to be able to redefine Wi-Fi performance in the world’s most demanding venues such as football grounds.</p> <p>Putting the launch into context, Extreme stressed that in the sporting arena, wireless communication is no longer a convenience layer but has instead become a core part of venue operations and how the fan experience is delivered.</p> <p>The company noted that <a href="https://www.computerweekly.com/news/366642758/Extreme-Connect-26-Wi-Fi-7-line-aims-to-address-needs-of-6GHz-era">modern stadiums and large venues demand more from wireless networks</a> than ever before, in particular fast, reliable connectivity for mobile ticketing, concessions, streaming, access to real time team/player stats, and social sharing. And at the same time as such fans’ needs are being met, venue operators are depending on wireless connectivity to support staff communications, operations, security, analytics, and revenue generating digital experiences without Wi-Fi dead zones.</p> <p>Yet, traditional <a href="https://www.computerweekly.com/news/366626580/Aviva-Stadium-upgrades-5G-infrastructure">stadium Wi-Fi designs</a> have often relied on omni-directional and directional coverage as well as dense under-seat deployment models to work around RF constraints. Yet while these approaches can deliver strong results, Extreme said they can also increase infrastructure mounting points, cable management requirements and long-term operational burden.</p> <p>To address these issues, Extreme claims that <a href="https://www.extremenetworks.com/resources/solution-brief/extreme-multi-beam-wireless">Extreme Multi-Beam Wireless</a> can deliver “unprecedented” capacity and precision Wi‑Fi performance for stadiums and large venues using sectorised, overhead coverage.</p> <p>Fundamentally, this is said to allow stadium owners to deliver targeted, high‑density Wi‑Fi using 16 independently, extending flexibility with a complete Wi‑Fi 7 portfolio for bowl, concourse and outdoor coverage. Built for high-density environments, it is engineered to increase capacity while supporting more devices and higher bandwidth demands with less complexity and “dramatically” less infrastructure, driving substantial capital and operational benefits for venue operators.</p> <p>It also designed to allow for serviced RF sectors, aligning capacity directly to user density for improved fan and staff experiences. In addition, the technology is attributed with reducing interference with defined coverage zones and controlled RF patterns and simplifying deployments by consolidating infrastructure into centralised overhead systems.</p> <p>The technology’s multi-beam architecture supports extending high-performance coverage into traditionally hard-to-reach areas, delivered through the strategic relationship with <a href="https://www.matsing.com/">MatSing</a>. This is said to result in the industry's first 16-sector directional antenna system, combining MatSing's lens antenna technology with sixteen Extreme AP5022FX <a href="https://bit.ly/ap5022">Wi-Fi 7 access points</a> in a one-to-one AP-to-sector architecture.</p> <section class="section main-article-chapter" data-menu-title="Targeted overhead coverage"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Targeted overhead coverage</h2> <p>Rather than treating a stadium bowl as a single coverage area, each system divides the environment into 16 independently serviced RF sectors, delivering targeted overhead coverage, predictable RF boundaries and capacity aligned to user density. By deploying multiple systems above the stadium bowl, Extreme and MatSing say operators can extend coverage and maintain consistent performance across large seating areas from a centralised deployment model.</p> <p>“We’ve created a first-of-its-kind solution that fundamentally changes how stadiums deliver connectivity,” commented MatSing CEO Bo Larsson. “Together, we’re enabling venues to support more fans, more devices, and more data than ever before, while delivering greater efficiency, reliability, and performance for the ultimate fan experience.”</p> <p>“Every stadium and arena presents unique wireless challenges, so venue operators need more than a one-size-fits-all Wi-Fi design,” added David Coleman, director of wireless in the office of the CTO at Extreme Networks. “Alongside our omni-directional, directional, and under-seat access points, the solution gives venues the flexibility to optimise coverage, maximise capacity and deliver exceptional fan experiences.”</p> <p>Extreme Multi-Beam Wireless will be available in the fourth quarter of the 2026 calendar year with the <a href="https://www.tennesseetitans.com/">Tennessee Titans</a> being the first to deploy the technology at its Nissan Stadium, opening Spring ’27.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about stadium connectivity</h3> <p><a href="https://www.computerweekly.com/news/366642758/Extreme-Connect-26-Wi-Fi-7-line-aims-to-address-needs-of-6GHz-era">Extreme Connect 26: Wi-Fi 7 line aims to address needs of 6GHz era</a>: Wi-Fi 7 access points designed to provide reliable, high-speed connectivity and key deployment made at University of Florida arena.</p> <p><a href="https://www.computerweekly.com/news/366630109/Neos-Networks-to-deliver-live-broadcast-connectivity-for-Premier-League">Neos Networks to deliver live broadcast connectivity for Premier League</a>: Provider of global connectivity services for broadcast and events taps dedicated B2B network provider to ensure high-quality delivery of Premier League football.</p> <p><a href="https://www.computerweekly.com/news/366618530/EE-scores-Wembley-Stadium-connectivity-renewal">EE scores Wembley Stadium connectivity renewal</a>:&nbsp;&nbsp;UK’s leading mobile comms provider extends title connectivity partnership with England’s national football stadium in London, as well as with all four Home Nations Football Associations.</p> <p><a href="https://www.computerweekly.com/news/366637319/Chelsea-shoots-for-enhanced-mobile-connectivity-in-and-around-Stamford-Bridge">Chelsea shoots for enhanced mobile connectivity in and around Stamford Bridge</a>: Leading English football club upgrades mobile network in and around home stadium to deliver enhanced connectivity for fans attending matches and events, based on new and upgraded small cells.</p> </div> </div> <p></p> <p>In a separate deployment to ensure high quality coverage for fans and operations, RF ball-shaped lens antennas from MatSing are being deployed in or around&nbsp;15 out of the 16 host World Cup venues&nbsp;in Mexico, Canada and the US. This includes all 11 NFL stadiums hosting World Cup matches in the US such as AT&amp;T Stadium in Dallas,&nbsp;Levi's Stadium in Santa Clara,&nbsp;Atlanta’s Mercedes-Benz Stadium, the MetLife Stadium in New Jersey and the SoFi Stadium in Los Angeles.</p> <p>The antennas - described as “unique” support 5G/4G connectivity for tens of thousands of spectators attending each match,&nbsp;optimising spectrum efficiency and providing coverage throughout the seating bowl, concourses and suites.&nbsp;Unlike traditional sector antennas, MatSing says the spherical lens approach delivers highly focused beams, with the result of enabling operators to “dramatically” increase capacity, reduce interference and simplify network design in some of the most challenging RF environments.</p> <p></p> <p></p> </section> Strategic partnership delivers stadium connectivity promising increased capacity, simplified operations and reduced infrastructure requirements https://cdn.ttgtmedia.com/rms/computerweekly/AT&T-stadium-joe-ohalloran.jpg https://www.computerweekly.com/news/366644929/World-Cup-venues-score-with-Extreme-MatSing-multi-beam-Wi-Fi Fri, 26 Jun 2026 10:30:00 GMT <p>With artificial intelligence (AI) firmly established in its product offerings, global comms tech provider Nokia has elevated its position in AI-ready networking, expanding its relationship with Amazon Web Services (AWS) to add its Autonomous Network Fabric to the hyperscaler. It also completed a joint proof of concept (PoC) with Databricks, demonstrating a unified, substrate-agnostic data platform designed to support AI-driven autonomous networks.</p> <p>With availability expected later in 2026, the collaboration adds Nokia’s <a href="https://www.computerweekly.com/news/366626110/Nokia-launches-autonomous-network-fabric">Autonomous Network Fabric</a> to AWS with the core intention of addressing the challenge of networks historically relying on siloed operational and business support systems.</p> <p>The Autonomous Network Fabric’s core capabilities include : unified data management across domains, agentic AI for service operations and optimisation, and <a href="https://www.computerweekly.com/news/366637819/Forward-Networks-claims-first-network-digital-twin-for-enterprises">digital twin</a> simulations for proactive impact assessment.</p> <p>For AWS, this offers <a href="https://www.computerweekly.com/feature/How-AI-is-being-used-to-manage-networks">network observability,</a> analytics, security and automation together through unified data management, agentic AI for service operations and optimisation, digital twin simulations and intent-based networking.</p> <p>Businesses will now have access to advanced AI and cloud services required for Level 4 autonomy. This builds on a set of existing digital operations applications from Nokia – covering orchestration, assurance and unified inventory – already on the platform.</p> <p>The broad Nokia portfolio will now deliver intent-based service orchestration across multi-domain, multi-supplier networks, and provides 360-degree observability with AI-powered anomaly detection, root cause analysis and closed-loop resolution. Importantly, it looks to offer a single source for network topology and resources.</p> <p>Running on AWS, Nokia said its solutions will gain elastic scalability, global availability and broad model choices through cloud, AI and machine learning (ML) services – including Amazon Bedrock and <a href="https://aws.amazon.com/sagemaker/ai/">Amazon SageMaker</a> – enabling operators to innovate faster while reducing infrastructure costs.</p> <p>The net result is said to be intent-based networking that translates business goals into automated closed-loop actions. Nokia is also engineering an optimised cloud footprint designed to that minimise compute and storage requirements versus traditional on-premise deployments.</p> <p>Meanwhile, the PoC with data and AI company <a href="https://www.databricks.com/">Databricks</a> aimed to show how network providers can simplify fragmented data environments and deploy real-time analytics at scale, enabling faster decision-making, improved network performance and more efficient operations.</p> <p>The PoC looked to address a long-standing industry challenge: dealing with various siloed operational and business support systems, each with its own data architecture, making it difficult to apply AI consistently across domains. Nokia stressed that to truly harness AI and multi-agent systems, operators needed a common data platform that can run seamlessly across different cloud environments or on-premise infrastructure, without the need to rewrite code.</p> <p>The POC showed the ability to develop a joint architecture that handles the massive scale and real-time ingestion speeds required to feed network data to AI agents for automated, cross-domain decision-making.</p> <p>Engineering teams from both countries focused on a real-time performance management use case, simulating analytics ingestion with an intent to scale quickly to match tier-1 operator scale in the cloud.</p> <p>Their work is said to have delivered several key technical breakthroughs designed to simplify how telecom operators build and run data-driven services across different environments cross-platform data pipelines, without coding complexity; supplier-neutral data logic design; automated deployment across environments; AI-powered creation of new data products; and a data fabric built for the agentic world.</p> <p>Data pipelines were created once and deployed across different platforms without modification. In trials, the same data workflows ran on both Databricks and an open source stack based on Apache Flink, Kafka and Iceberg, supporting real-time streaming, batch processing and query-time data products.</p> <p>To avoid lock-in to any single platform, Nokia engineers developed transformation logic using an abstract, platform-independent expression in Python. By separating the core logic from platform-specific connectors, the same data workflows were reused across multiple environments.</p> <p>The teams also validated a custom compiler that automatically adapted workflows at deployment. Based on the target environment, it translated the abstract logic into native formats and added the platform-specific connectors, eliminating manual rework and accelerating time to deployment.</p> <p>Using simple natural language prompts, an intelligent data fabric agent generated new data products, requested human validation and deploy the pipeline automatically, resulting in faster innovation with less manual effort. In the agentic world, the same mechanism can be used by other agents to create dynamic data products on demand by communicating agentic with the data fabric agent.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI in networking</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/feature/How-AI-is-being-used-to-manage-networks">How AI is being used to manage networks</a>: Network management is becoming reliant on artificial intelligence-enabled tools, which use machine learning based on network monitoring data.</li> <li><a href="https://www.computerweekly.com/news/366644358/Cisco-36-months-to-modernise-networks-before-AI-overwhelms-capacity">36 months to modernise networks before AI overwhelms capacity</a>: Research finds capacity and performance the top network challenge for UK organisations, with 81% of respondents saying their network does not have room to house evolving AI demands.</li> <li><a href="https://www.computerweekly.com/news/366642566/Extreme-Connect-26-Agent-ONE-takes-forward-network-AI">Extreme Connect 26: Agent ONE takes forward network AI</a>: Network firm launches ‘smarter, faster, autonomous’ approach to enterprise networking, with its operating model moving from assistive AI to autonomous, always-on operations.</li> <li><a href="https://www.computerweekly.com/news/366639478/Nokia-AWS-demo-agentic-AI-network-slicing-with-du-Orange">Nokia, AWS&nbsp;demo agentic AI network slicing with Du, Orange</a>: ‘Industry-first’ intent-based 5G-Advanced slicing with agentic AI to offer providers with premium network slicing services that respond to real-world situations and enable autonomous intelligence.</li> </ul> </div> </div> Spate of activity sees global comms tech provider announce expanded collaboration with hyperscaler, as well as a joint proof of concept with data and AI company to support autonomous networks for the AI era https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/wireless-network-mobile-broadband-1-adobe.jpeg https://www.computerweekly.com/news/366645100/Nokia-accelerates-AI-for-networking-drive Thu, 25 Jun 2026 10:38:00 GMT <p>Qualcomm has reached an agreement to acquire artificial intelligence (AI)-native software stack provider <a href="https://www.modular.com/">Modular</a>.</p> <p>Modular provides an open, AI-native software stack that enables artificial intelligence to run efficiently across hardware architectures. The company’s unified platform runs models across central processing unit, graphics processing unit, neural processing unit and custom application-specific integrated circuit architectures without rewrites for each accelerator.</p> <p>For developers and enterprises, this is said to mean building once, deploying across any environment with lower total cost of ownership. Modular also believes it is supported by an open, industry-friendly, supplier-neutral developer community committed to improving the portability and efficiency of AI infrastructure.&nbsp;</p> <p>Explaining the reason for its acquisition, <a href="https://www.computerweekly.com/news/366641586/Qualcomm-expands-strategic-advanced-driver-assistance-systems-immersive-eyewear-collaborations">connected comms tech platform provider</a> Qualcomm said addressing the pressures and demands faced by companies requires more than hardware. It added that developers need software that connects system-level optimisation with heterogeneous, disaggregated compute, turning silicon performance into reliable and efficient AI services across accelerators, environments and use cases.</p> <p>Qualcomm believes the acquisition will provide a software foundation for generative and agentic AI across datacentre and edge environments supporting <a href="https://www.computerweekly.com/news/366633526/Qualcomm-gears-up-for-AI-inference-revolution">more efficient inference</a>. This datacentre strategy encompasses orchestration and deployment in distributed AI systems, while strengthening relationships with model creators, developers, hyperscalers and enterprises.</p> <p>Qualcomm also expects the deal to strengthen its Technologies division’s ability to deliver a more optimised AI compute layer across a broad range of platforms and use cases.</p> <p>Specifically, by combining what it claims is its own silicon leadership with Modular’s software expertise, Qualcomm insists the Technologies division will be well positioned to help customers move AI into production from device to cloud, with systems that are faster, more efficient and easier to scale.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about AI in networking</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/feature/How-AI-is-being-used-to-manage-networks">How AI is being used to manage networks</a>: Network management is becoming reliant on artificial intelligence-enabled tools, which use machine learning based on network monitoring data.</li> <li><a href="https://www.computerweekly.com/news/366644358/Cisco-36-months-to-modernise-networks-before-AI-overwhelms-capacity">36 months to modernise networks before AI overwhelms capacity</a>: Research finds capacity and performance the top network challenge for UK organisations, with 81% of respondents saying their network does not have room to house evolving AI demands.</li> <li><a href="https://www.computerweekly.com/news/366642566/Extreme-Connect-26-Agent-ONE-takes-forward-network-AI">Extreme Connect 26: Agent ONE takes forward network AI</a>: Network firm launches ‘smarter, faster, autonomous’ approach to enterprise networking, with its operating model moving from assistive AI to autonomous, always-on operations.</li> <li><a href="https://www.computerweekly.com/news/366639478/Nokia-AWS-demo-agentic-AI-network-slicing-with-du-Orange">Nokia, AWS&nbsp;demo agentic AI network slicing with Du, Orange</a>: ‘Industry-first’ intent-based 5G-Advanced slicing with agentic AI to offer providers with premium network slicing services that respond to real-world situations and enable autonomous intelligence.</li> </ul> </div> </div> <p><a href="https://www.computerweekly.com/news/366566053/CES-2024-automotive-to-the-fore-as-Qualcomm-revs-up-for-new-world-of-AI">In a keynote session at&nbsp;the CES 2024 event</a>, Qualcomm CEO Cristiano Amon observed that the overarching strategy for the company would be to tap into the growing world of AI in local devices. He added that the company was undergoing a fundamental shift in how it sees itself, from a wireless communication company to being a connected processor and artificial intelligence firm.</p> <p>Fast forward two years, and describing the latest part of this strategy, Amon said the Modular acquisition marks a pivotal moment not just for the company, but for the AI industry at large.</p> <p>“As agentic AI <a href="https://www.computerweekly.com/news/366634654/Cisco-Network-architectures-must-be-rebuilt-for-agentic-AI">scales across datacentres and edge environments</a>, the industry is moving toward disaggregated, multi-vendor architectures that demand a more open and modern software foundation,” he said.</p> <p>“We believe the future belongs to developer-friendly, horizontal platforms that can run across diverse compute environments and give customers real choice in how and where they deploy AI. With Modular, we’re accelerating that shift, combining our scale and energy-efficient datacentre technologies with an open ecosystem approach to help drive the next chapter of AI.”</p> <p>Modular co-founder and CEO Chris Lattner said: “Modular was founded on the belief that AI needs a more open and efficient software foundation that can span diverse hardware and deployment environments.</p> <p>“Joining Qualcomm gives us the scale and platform reach to accelerate that mission,” he added. “Together, we can make AI development more accessible and performant for developers, strengthen portability across hardware, and help grow an open ecosystem that broadens participation and speeds innovation. We are excited to continue advancing our software platform as part of Qualcomm’s broader strategy from edge to cloud.”</p> <p>The transaction is expected to close in the second half of 2026, subject to customary closing conditions and applicable regulatory approvals.</p> Connected comms tech platform provider makes strategic buyout to accelerate adoption of edge-to-cloud AI platforms by developers, OEMs, ODMs, cloud service providers and model creators https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/Qualcomm-Control-Center-hero.jpg https://www.computerweekly.com/news/366645173/Qualcomm-gains-AI-edge-with-Modular-acquisition Thu, 25 Jun 2026 07:45:00 GMT <p>Stating that it can deliver enterprise satellite internet deployed in days, not months, IT solutions provider WEI has launched a managed enterprise connectivity solution powered by SpaceX’s Starlink service.</p> <p>WEI has said that the <a target="_blank" href="http://www.connect.wei.com/" rel="noopener">WEI Connect</a> system was designed for enterprise expectations, delivering enterprise-ready, always-on Starlink connectivity that lets organisations move at the speed of their business. It stated that the system will give IT leaders a new level of control over their networks, whether they’re connecting remote offices, ensuring business continuity or enabling modern wide area networks (WAN) strategies.</p> <p>WEI Connect aims to enable organisations to deploy resilient, high-speed internet at fixed or mobile sites in hours rather than the 60 to 180 days typically required for <a href="https://www.computerweekly.com/news/366642948/Zayo-Europe-opens-Genoa-fibre-network-landing-interconnection-hub">traditional fibre</a> or coax circuits. The system looks to maintain continuity across fixed or mobile sites to reduce business disruptions with “enterprise-grade” solutions that integrate into existing infrastructure, simplifying operations.</p> <p>As part of the overall solution offer, WEI is offering <a href="https://starlink.com/gb/business">Starlink’s Mini, Standard and Performance user terminals</a>, tailored for organisations that require high-throughput, low-latency internet access at fixed or mobile sites. WEI claimed that its Connect Starlink user terminal can be installed and activated in a matter of hours, with preconfigured deployment kits and local installation services enabling enterprises to bring new sites online rapidly.</p> <p>WEI said that <a href="https://www.computerweekly.com/news/366642501/TMobile-Starlink-aim-to-reinvent-business-internet-from-ground-up-sky-down">Starlink is “changing what’s possible” for enterprise connectivity</a>, but that its real value comes from how it’s deployed, supported and integrated. To make Starlink enterprise-ready, WEI Connect will provide total visibility through a single platform and partner.</p> <p>While Starlink provides hardware powerful enough to support enterprise networking use cases, WEI stressed that it takes an experienced IT partner to make it truly enterprise-ready and that the company could offer 24x7 support to resolve issues ahead of evolving requirements.</p> <p>Overall, WEI Connect with Starlink is claimed to deliver on four core enterprise outcomes: built-in resiliency, ready in days, not months; simplified operations with centralised visibility; cost-effective primary and backup connectivity; 24x7 expert support and rapid deployment at scale.</p> <p>In the former regard, preconfigured rapid deployment kits can be installed by local technicians, and WEI validates and integrates Starlink with leading S<a href="https://www.computerweekly.com/news/366635495/SASE-SD-WAN-evolve-as-enterprises-prioritise-unified-network-security">D-WAN and SASE platforms</a> including <a href="https://www.computerweekly.com/news/366624962/HPE-Aruba-Networking-expands-distributed-services-switch-platform">HPE Aruba Networking</a>, Palo Alto Networks, Fortinet, <a href="https://www.computerweekly.com/news/366644358/Cisco-36-months-to-modernise-networks-before-AI-overwhelms-capacity">Cisco</a> and Cato Networks.</p> <p>The system integrates into existing stacks, making it easy to manage distributed sites centrally and consistently. A purpose-built dashboard looks to deliver real-time visibility into performance, usage, costs, site health and service-level agreement (SLA) insights. Customised and pooled data plans can be tailored for a site’s unique needs.</p> <p>Primary and backup internet solutions with custom consumption models can be set to align with standard procurement models and customise billing while hands-on, high-touch customer service with stocked inventory is ready for deployment and device replacement.</p> <p>WEI said that it has more than 4,000 certified field technicians across North America, dedicated spare hardware and SLAs offering four-hour response or next-business-day replacement, meaning IT leaders can be “confident that downtime will be minimised”.</p> <p>Among other key capabilities, WEI Connect integrates directly with the Starlink API to deliver a centralised performance dashboard. This generates real-time telemetry across the enterprise, including usage tracking, billing by site or region, link speed, throughput, latency and signal quality. It also delivers alerts for alignment or obstructions, thermal conditions, power disconnects and pending software reboots, along with custom reporting and helpdesk ticketing integration.</p> <p>Commenting on the utility of the combined system, Greg LaBrie, vice-president and general manager at WEI, said: “WEI Connect [can adapt] to the changing needs of your operations and delivering the reliability and expertise modern organisations require. With WEI’s engineering expertise, managed services and field deployment capabilities, enterprises can confidently adopt Starlink as part of a resilient, secure networking strategy.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about space-based communications</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366641709/Amazon-acquires-Globalstar-to-expand-satellite-comms-business">Amazon acquires Globalstar to expand satellite comms business</a>: Strategic purchase to see satellites, radio frequency spectrum and operational expertise to enable existing Leo business to add direct-to-device services to future generations of its low Earth orbit satellite offer.</li> <li><a href="https://www.computerweekly.com/news/366643328/Viasat-hosts-first-satellite-enabled-phone-comms-in-Uzbekistan">Viasat hosts first satellite-enabled phone comms in Uzbekistan</a>: Satellite communications successfully demonstrates direct-to-device satellite connectivity in Uzbekistan for the first time, looking to position the country as a key market pioneering the progression of satellite technology.</li> <li><a href="https://www.computerweekly.com/news/366643796/Direct-to-cell-growth-hits-headwinds-while-6G-set-for-rapid-uptake">Direct-to-cell growth hits headwinds while 6G set for rapid uptake</a>: Research predicts monthly active satcom users to reach over 130 million by 2031, but usage forecast to be lower than anticipated, while 6G services set to grow from 4.6 million in 2029 to 2.9 billion in 2035.</li> <li><a href="https://www.computerweekly.com/news/366641075/SpaceLocker-launches-first-shared-satellite-mission">SpaceLocker launches first shared satellite mission</a>: French startup looks to redefine space infrastructure by turning satellites into shared platforms, and claims faster, more sustainable missions.</li> </ul> </div> </div> Managed connectivity solution pairs leading satellite network with enterprise integration and field technicians to deliver resilient primary and backup internet across every site https://cdn.ttgtmedia.com/visuals/LeMagIT/hero_article/Hero-starlink.jpg https://www.computerweekly.com/news/366644770/WEI-intros-Starlink-powered-managed-enterprise-connectivity Wed, 24 Jun 2026 12:40:00 GMT <p>Artificial intelligence has fundamentally altered <a href="https://www.computerweekly.com/resources/Data-centre-energy-efficiency-and-green-IT">the sustainability conversation</a> within enterprise IT.</p> <p>For years, organisations made steady progress in improving the efficiency of their digital estates – consolidating workloads, migrating to cloud platforms and <a href="https://www.computerweekly.com/opinion/IT-Sustainability-Think-Tank-The-role-of-procurement-in-closing-the-sustainability-gap">embedding sustainability into procurement</a> and reporting frameworks. Those efforts, while meaningful, were largely built around a predictable model of demand.</p> <p>AI changes that model entirely.</p> <p>High-density compute is no longer optional. It is becoming a core requirement for competitiveness, innovation and in some cases, operational survival. The challenge for CIOs is not whether to embrace it, but how to do so without undermining the sustainability commitments many organisations have spent years establishing.</p> <p>The reality is that traditional approaches to sustainability target setting are no longer sufficient.</p> <p>Targets must now be achievable, measurable and, critically, grounded in operational reality. Otherwise, there is a risk they become detached from the infrastructure strategies required to deliver business value.</p> <h3>Moving from ambition to operationally-achievable targets</h3> <p>One of the most common pitfalls in sustainability strategy is setting targets that look credible on paper but are disconnected from how technology is actually deployed and consumed.</p> <p><a href="https://www.computerweekly.com/opinion/Realities-of-the-AI-age-force-sustainability-to-the-fore">In an AI-driven environment</a>, this gap becomes more pronounced.</p> <p>CIOs need to move away from broad top-down commitments and instead define targets that are embedded within infrastructure decision-making. That means aligning <a href="https://www.computerweekly.com/opinion/IT-Sustainability-Think-Tank-The-importance-of-measurement-in-closing-the-sustainability-gap">sustainability metrics</a> directly to workload design, data management and hardware lifecycle planning.</p> <p>For example:</p> <ul class="default-list"> <li> <p>Defining acceptable energy intensity thresholds for AI workloads, rather than treating all compute equally&nbsp;</p> </li> <li> <p>Establishing clear policies on model training frequency and dataset retention&nbsp;</p> </li> <li> <p>Embedding lifecycle extension targets for physical infrastructure alongside performance objectives.</p> </li> </ul> <p>These are not headline-grabbing commitments, but are achievable, enforceable and capable of being audited.</p> <p>Sustainability, in this context, becomes less about aspiration and more about engineering discipline.</p> <h3>Beyond market-based reporting</h3> <p>A second challenge lies in how sustainability performance is measured and reported.</p> <p>Many organisations continue to rely heavily on market-based carbon accounting, supported by renewable energy certificates and offset mechanisms. While these have a role to play, they can create a misleading picture of actual environmental impact.</p> <p>The shift towards location-based reporting is therefore essential.</p> <p>Understanding where workloads run, how energy is generated in those locations and how grid intensity fluctuates over time provides a far more accurate reflection of environmental impact. It also enables more informed decision-making at an architectural level.</p> <p>However, this requires greater transparency than many organisations currently have access to.</p> <p>As highlighted in earlier discussions around cloud sustainability, provider-level reporting often lacks the granularity required for meaningful enterprise analysis. Without consistent methodologies and comparable data, CIOs are left working with approximations rather than auditable metrics.</p> <p>To address this, organisations need to combine external data with internal governance:</p> <ul class="default-list"> <li> <p>Correlating workload placement with regional carbon intensity data&nbsp;</p> </li> <li> <p>Building internal reporting frameworks that standardise measurement across environments&nbsp;</p> </li> <li> <p>Challenging suppliers to provide more granular, verifiable data.&nbsp;</p> </li> </ul> <p>Only then can sustainability targets move from indicative to defensible.</p> <h3>Rethinking the AI refresh cycle</h3> <p>Perhaps the most significant, and least discussed, sustainability risk associated with AI is the potential for accelerated hardware refresh cycles.</p> <p>The performance demands of AI workloads are driving rapid adoption of specialised infrastructure, particularly GPU-intensive environments. While this delivers clear capability gains, it also creates a temptation to prematurely retire existing assets in favour of new, optimised platforms.</p> <p>This is where sustainability strategy must take a more balanced view.</p> <p>The embodied carbon associated with manufacturing new hardware is substantial. In many cases, the environmental cost of early replacement outweighs the operational efficiency gains delivered by newer equipment.</p> <p>Extending the life of legacy infrastructure, where appropriate, therefore becomes a critical lever.</p> <p>This does not mean resisting innovation or compromising performance. It means adopting a more nuanced approach:</p> <ul class="default-list"> <li> <p>Segregating workloads so that high-density AI compute runs on optimised platforms, while less intensive tasks remain on existing infrastructure&nbsp;</p> </li> <li> <p>Identifying opportunities for redeployment rather than wholesale replacement&nbsp;</p> </li> <li> <p>Integrating lifecycle extension and transition planning into procurement and refresh strategies.&nbsp;</p> </li> </ul> <p>Crucially, organisations also need to consider what happens at the point of transition.</p> <p>Decisions made at end-of-life – whether assets are redeployed, reused, or prematurely retired – have a direct and often underappreciated impact on overall sustainability performance. In many cases, these moments represent one of the few points in the infrastructure lifecycle where outcomes can be fully measured, verified and audited, rather than inferred.</p> <p>Ignoring this stage risks undermining otherwise well-intentioned sustainability strategies.</p> <h3>Sustainability as a differentiator</h3> <p>While much of the sustainability conversation is framed in terms of risk mitigation or compliance, there is a growing opportunity for organisations to use it as a genuine differentiator.</p> <p>This is particularly true in sectors where clients, regulators and investors are placing increasing emphasis on verifiable environmental performance.</p> <p>The key word here is verifiable.</p> <p>Organisations that can demonstrate the following will be in a far stronger position than those relying on high-level claims or offset-driven narratives:</p> <ul class="default-list"> <li> <p>Clear alignment between infrastructure strategy and sustainability targets&nbsp;</p> </li> <li> <p>Transparent, auditable reporting methodologies&nbsp;</p> </li> <li> <p>Responsible management of technology across its full lifecycle, including how assets are transitioned, redeployed and retired.</p> </li> </ul> <p>In practice, this often comes down to control.</p> <p>Enterprises may have limited visibility into upstream infrastructure operated by hyperscale providers, but they retain direct control over how their own technology estate is managed, particularly at points of refresh, redeployment and end-of-life.</p> <p>Those control points provide a tangible foundation for building sustainability strategies that are not only credible, but defensible under scrutiny.</p> <p>In an environment where AI adoption is accelerating, this level of accountability becomes a meaningful differentiator.</p> <h3>A shift in accountability</h3> <p>Ultimately, the move towards sustainable AI infrastructure requires a shift in how responsibility is understood.</p> <p>It is no longer sufficient to view sustainability as a function of the datacentre operator or cloud provider alone. Enterprises themselves are active participants in driving demand and shaping outcomes.</p> <p>As discussed in the context of AI infrastructure more broadly, environmental impact is the cumulative result of countless individual decisions, from workload design to data retention to hardware refresh cycles.</p> <p>Importantly, some of the most impactful of these decisions occur at transition points within the lifecycle.</p> <p>How long assets are retained, how effectively they are redeployed, and how they are ultimately retired are not peripheral considerations. They are central to whether sustainability targets can be realistically achieved and evidenced.</p> <p>These are also areas where organisations have the greatest degree of control.</p> <p>CIOs therefore have a critical role to play.</p> <p>Not in limiting innovation, but in ensuring that innovation is delivered with a full understanding of its implications. Not just in production, but across the entire lifecycle of the technology that enables it.</p> <h3>Conclusion</h3> <p>The tension between AI adoption and sustainability is real, but it is not insurmountable.</p> <p>By focusing on achievable, operationally-grounded targets, moving towards more accurate and transparent reporting, and taking a lifecycle view of infrastructure, organisations can navigate this challenge effectively.</p> <p>In doing so, they not only protect their sustainability commitments, but create an opportunity to differentiate.</p> <p>Because in an AI-driven world, it will not be enough to demonstrate what your infrastructure can do.</p> <p>Increasingly, organisations will also be judged on how responsibly they choose to run it.</p> <p>For organisations looking to strengthen this aspect of their strategy, aligning infrastructure decisions with robust <a href="https://www.astralistech.co.uk/itad">secure IT asset disposal</a> practices can provide a practical foundation for achieving auditable sustainability outcomes.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about IT and sustainability</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/opinion/How-to-improve-AI-efficiency-beyond-cost-optimisation">How to improve AI efficiency beyond cost optimisation</a>. With half of generative AI projects expected to overrun budgets by 2028, IT leaders must drive efficiency across the AI stack to protect margins and address environmental challenges</li> <li><a href="https://www.computerweekly.com/opinion/What-you-need-to-know-before-emissions-regulators-come-knocking">What you need to know before emissions regulators come knocking</a>. Carbon emissions reporting is becoming mandatory. But accounting is not the same as reducing, especially given the smoke and mirrors in some carbon footprint reporting</li> </ul> </div> </div> <p></p> <p></p> AI demands high-density compute, challenging sustainability goals. CIOs must shift from vague targets to operational, audited metrics and responsible hardware lifecycle management https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/IT-sustainability-think-tank-hero.jpg https://www.computerweekly.com/opinion/Setting-achievable-sustainability-targets-in-the-age-of-AI-infrastructure Wed, 24 Jun 2026 06:55:00 GMT <p>In two separate Executive Orders (EOs) signed on 22 June by US president Donald Trump, the US government has been instructed to take action to establish a more cohesive approach to the commercialisation and deployment of <a href="https://www.whitehouse.gov/presidential-actions/2026/06/ushering-in-the-next-frontier-of-quantum-innovation/" target="_blank" rel="noopener">quantum computing</a>, sensing and networking, and to work to <a href="https://www.whitehouse.gov/presidential-actions/2026/06/securing-the-nation-against-advanced-cryptographic-attacks/" target="_blank" rel="noopener">mitigate future cyber security threats</a> arising from quantum computers.</p> <p>The EO covering quantum information science and technology (QIST), builds on the <a href="https://www.congress.gov/bill/115th-congress/house-bill/6227" target="_blank" rel="noopener">National Quantum Initiative Act</a> signed in 2018, during the president’s first term in office.</p> <p>Setting out Trump’s desire for the US to maintain a strategic technological advantage in this field, the EO gives the government six months to update America’s National Quantum Strategy to support the fast-maturing ecosystem and promote commercialisation and deployment of quantum technology.</p> <p>It also establishes the Quantum Computer for Application Development and Discovery Science (QC-ADDS) Effort – a cross-body programme – to pursue a quantum computer at a scale that will “initiate the era of quantum-enabled scientific discovery”. Under the proposed terms of this programme, at least one such computer is to be delivered to the Department of Energy, and if possible, opened up to the wider scientific community.</p> <p>Additionally, the US government will work to identify at least three next-gen quantum sensor projects to prioritise by the end of September 2028, and to develop a plan to advance the commercial readiness of quantum sensing, sensor-manufacturing technology, and quantum network-enhanced timing.</p> <p>Meanwhile, Washington will also begin work on a plan to strengthen the wider QIST ecosystem, supporting supply chains, private sector adoption of QIST standards, and R&amp;D paths, while another project will aim to ensure cyber security concerns around QIST are allayed, and balanced controls implemented that protect US interests without stifling innovation, as teams from the State and Commerce departments work to align with international allies and partners such as the UK.</p> <p>The second order sets out to reinforce cyptographic security for the US’s sensitive data, CNI and digital economy, and looks to effectively execute the transition of government IT systems to post-quantum cryptography (PQC) standards <a href="https://www.nist.gov/pqc" target="_blank" rel="noopener">as set out by the National Institute for Standards and Technology</a> (NIST), and to assist CNI owners and operators as they conduct their own upgrades.</p> <section class="section main-article-chapter" data-menu-title="Quantum is no longer just a scientific issue"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Quantum is no longer just a scientific issue</h2> <p>Carlos Moreira, chair and CEO of <a href="https://www.sealsq.com/" target="_blank" rel="noopener">Sealsq</a>, which is working to develop post-quantum security technology, reflected the thinking of the wider IT community when he said the EO acknowledged that quantum has now moved from a purely scientific conversation to one of security and economic competitiveness.</p> <p>“As governments invest heavily in quantum computing, protecting research data, intellectual property and supply chains from cyber espionage becomes increasingly important,” he said.</p> <p>“One of the key challenges is that sensitive information stolen today could be stored and decrypted in the future as quantum capabilities mature. That makes quantum-resistant security measures a consideration now, particularly for organisations involved in advanced research, semiconductors, AI and critical infrastructure.</p> <p>“The broader trend is that securing the quantum ecosystem will require more than protecting research labs,” said Moreira. “It will also mean safeguarding the devices, networks, software and supply chains that support quantum innovation.”</p> <p>Stefan Leichenauer, vice-president of engineering at <a href="https://www.sandboxaq.com/" target="_blank" rel="noopener">SandboxAQ</a>, an AI and quantum specialist spun out of Google parent Alphabet, said: “The US has a window of opportunity to lead in this domain. It requires coordinated investment across the stack: cryptography, compute infrastructure, data generation and application development. It also requires strong partnerships between government, industry and academia to focus our resources and talent towards a common goal.</p> <p>“The new White House Executive Order on quantum will provide the spark needed to stimulate quantum leadership and innovation in the years to come,” added Leichenauer.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about quantum computing</h3> <ul class="default-list"> <li>Quantum computing presents unique sustainability challenges due to its specialised infrastructure and energy demands,&nbsp;<a rel="noopener" target="_blank" href="https://www.techtarget.com/sustainability/feature/How-does-quantum-computing-affect-sustainability">while also offering potential efficiency gains</a>.</li> <li>Microsoft researchers have made a breakthrough in quantum reliability&nbsp;<a rel="noopener" target="_blank" href="https://www.computerweekly.com/news/366643935/Agentic-AI-helps-Microsoft-speed-up-viable-quantum-computer">with the help of agentic artificial intelligence</a>.</li> <li>University of Waterloo proposes encrypted qubit encodings that enable one-time reconstruction,&nbsp;<a rel="noopener" target="_blank" href="https://www.techtarget.com/searchstorage/feature/Solving-quantum-computings-longstanding-no-cloning-problem">reframing quantum data resilience, security and storage</a>.</li> </ul> </div> </div> </section> In an Executive Order, president Trump directed the US government to work to establish a cohesive, collaborative approach to the development of quantum technology https://cdn.ttgtmedia.com/visuals/German/article/quantum-computing-3-adobe.jpg https://www.computerweekly.com/news/366645076/Trump-directs-US-government-focus-to-quantum Tue, 23 Jun 2026 14:45:00 GMT <p>For the past six weeks, Gulf and European banks have been war-gaming <a href="https://www.computerweekly.com/resources/Data-centre-capacity-planning">datacentre relocations</a>.&nbsp;</p> <p>Iranian drone and missile strikes on AWS facilities in the <a href="https://www.computerweekly.com/news/366644877/Gulf-CIOs-shift-focus-from-recovery-to-cyber-resilience-as-regional-threats-intensify">United Arab Emirates and Bahrain</a> in March caused damage that AWS itself now says could take several months to fully restore. UAE banks – including Abu Dhabi Commercial Bank, Emirates NBD and First Abu Dhabi Bank – are operating under weekly regulatory dispensations to use offshore datacentres because their primary cloud infrastructure cannot reliably serve them.</p> <p>Meanwhile, <a href="https://www.parallels.com/newsroom/news/press-releases/20260217-cloud-survey/">94% of IT leaders say they are seriously concerned about vendor lock-in</a>, and many are actively exploring exits.&nbsp;</p> <p>The hyperscalers have responded. IBM launched Sovereign Core in May. Microsoft has just scaled Azure Local to thousands of servers as a Sovereign Private Cloud configuration.&nbsp;</p> <p>Both products are pitched as <a href="https://www.computerweekly.com/feature/Is-cloud-data-sovereignty-all-just-a-case-of-Trust-me-bro">the answer to the sovereignty question</a>. They are not. They are a marketing fix to an architectural problem.</p> <p>The reason European governments keep ending up with these products is not that the underlying problem is unknown. It is that the psychology of large-scale IT procurement has not adjusted to the reality of geopolitical exposure.&nbsp;</p> <p>European politicians, in my experience working with them, are resolutely untechnical. The default psychology of public-sector cloud buying remains, “no one got fired for buying IBM”. So you get European ministers signing Google sovereign cloud contracts with clauses in them that are legally invalid against US law. And you get national sovereign cloud projects that are, when you look at the underlying stack, rebadged hyperscaler products.</p> <p>The Swiss <a href="https://www.computerweekly.com/feature/This-rise-of-the-splinternet-Data-sovereignty-risks-and-responses">federal sovereign cloud</a> is the cleanest example. The product being marketed as Switzerland's sovereign cloud is essentially rebranded Microsoft Azure. The Swiss military is refusing to use it. If you cannot get your own country's armed forces to deploy on the sovereign cloud you are pitching to your enterprises, you do not have a sovereignty product. You have a procurement artefact.</p> <p>But even where the underlying cloud is genuinely under your control, there is a second problem most of these procurement conversations have not caught up with yet. It is no good having a sovereign cloud running in your country that is open source and governed by you, but then using it to run American cloud software that has backdoors.&nbsp;</p> <p>So, if you create a sovereign Salesforce, take Salesforce as software, run it on your sovereign cloud, pay a licence fee, and call it air-gapped, you have the same problem one layer up. The application has backdoors.&nbsp;</p> <p>It probably also has kill switches. And kill switches stopped being hypothetical a long time ago. We got close to that with the Trump-Greenland episode last year, it is not inconceivable that an American administration would, under sufficient political pressure, start switching foreign deployments off.</p> <p>A sovereignty wrapper around hyperscaler infrastructure does not change where the underlying technology originates. It does not change the fact that the cloud software was written by an American or Chinese company, with whatever hidden backdoors and kill switches that come with that.&nbsp;</p> <p>And it does not change the calculus when the underlying infrastructure goes down. The UAE banks operating under weekly dispensations are not in that position because their sovereignty contracts failed. They are in that position because contracts cannot be invoked fast enough to move a banking workload off a damaged datacentre and onto another one in days, let alone hours.</p> <p>You can fund models, buy GPUs and launch national cloud initiatives. But if the infrastructure those systems run on is owned by foreign providers, you have not achieved sovereignty. You have created a dependency with better branding.</p> <p>The architectural fix looks different.&nbsp;</p> <p>Instead of putting a sovereign wrapper around someone else's stack, you build a cloud that is created by a mathematical network, combining compute nodes from independent parties, in different jurisdictions, governed by a protocol rather than a vendor. Your software then becomes immune to traditional infrastructure hacks, and you can change which providers run those nodes without interrupting your hosted software at all.</p> <p>The mechanism is straightforward. Say you have spread your cloud across Amazon datacentres, and you decide you want to move to Google. You add new Google nodes. Once they are synced into your subnet, you delete the Amazon nodes. Your apps and services continue running without a hitch. Your users do not notice anything. The cloud walks across the underlying compute. Your hosted software is completely independent of the underlying compute provider.</p> <p>The portability is mathematical. It is enforced at the protocol level using Byzantine fault tolerance, meaning a sufficient number of nodes have to agree on every state change for any state change to occur. If a hacker tries to move a workload, they cannot, because two plus two cannot equal five. Two plus two always equals four, and that is why it is hack-proof. You cannot have a Sybil-type attack on it either, because nodes have to submit to the governance system that runs on the network.</p> <p>That mathematical guarantee starts to look more attractive when the question shifts from regulatory exposure to kinetic risk. I have been showing the multi-provider, multi-jurisdiction model to Middle East clients specifically because of the drone attacks on datacentres.&nbsp;</p> <p>The traditional cloud resilience playbook, with multi-region failover and backup zones within the same provider, does not solve this problem. If a drone hits the datacentre your primary region is in, your failover region is statistically also at risk. The point of running the cloud as a mathematical network is that, even when the Iranians turn drones onto the datacentres, your stuff is still running on Amazon. It just does not go down.</p> <p>There is a reasonable counterargument that this architecture is hard. It is. It requires institutions to think differently about how compute is provisioned, governed and audited. It requires governments to accept that sovereignty is a property of the architecture, not of the contract. And it requires the wider industry to acknowledge that the hyperscaler-wrapper era, sovereign cloud branded by AWS, Azure or Google, sold as the answer to digital sovereignty, is a transition state, not a destination.</p> <p>But that transition is happening anyway. The next major outage at a hyperscaler datacentre, kinetic or otherwise, will force the conversation into a new register. The next signed-but-legally-invalid sovereign cloud contract will force a procurement reckoning. The institutions that survive both, and they are coming, on a timeline measured in months not years, will be the ones that designed for the architectural reality of multi-provider, multi-jurisdiction, mathematically-enforced portability long before they were asked to.</p> <p>Sovereignty is not a sticker you put on someone else's cloud. It is a property of how the system is built.</p> <p><a href="https://ch.linkedin.com/in/thedwilliams"><em>Dominic Williams</em></a><em> is founder of </em><a href="https://dfinity.org/"><em>Dfinity Foundation</em></a><em>, the not-for-profit research and development organisation behind the Internet Computer protocol.</em></p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about data sovereignty</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/feature/Is-cloud-data-sovereignty-all-just-a-case-of-Trust-me-bro">Is cloud data sovereignty all just a case of ‘Trust me, bro’?</a> Hyperscaler cloud is inherently global. Does that make data sovereignty unattainable – especially given the powers US courts hold? We grilled the hyperscalers in an attempt to find out</li> <li><a href="https://www.computerweekly.com/feature/Breaking-the-stranglehold-Responses-to-data-sovereignty-risk">Breaking the stranglehold: Responses to data sovereignty risk</a>. We look at the political and government responses to risks around data sovereignty and massive dependence on the three US hyperscalers – AWS, Azure and GCP – in the UK and Europe</li> </ul> </div> </div> <p></p> <p></p> Sovereign cloud wrappers fail against physical and legal risks. True sovereignty requires building mathematically-enforced, multi-jurisdictional infrastructure, not vendor contracts https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/drones-military-sorin-adobe.jpg https://www.computerweekly.com/opinion/Why-sovereign-cloud-is-a-marketing-fix-not-an-architectural-one Tue, 23 Jun 2026 06:33:00 GMT <p>nLighten’s <a href="https://www.computerweekly.com/news/366644836/nLighten-completes-15m-refurbishment-of-Bristol-edge-datacentre">£15m refurbishment of its Bristol datacentre</a> shows one section of the market placing its bets on relatively small edge infrastructure targeted at specific local tech-heavy customers.&nbsp;</p> <p>Following the launch of the Aztec West facility, we spoke to nLighten CEO Dawn Childs, who has a background working for the <a href="https://www.computerweekly.com/news/366639346/Demand-necessitates-digital-twin-and-data-visualisation-at-National-Grid">National Grid</a> and in datacentre development and operations. She is a strong champion of women and young people in engineering.&nbsp;</p> <p>She spoke to Computer Weekly about the practical realities of the UK market, the disconnect between the government’s artificial intelligence (AI) growth ambitions and energy availability in grid-constrained regions, data sovereignty concerns, and the <a href="https://www.computerweekly.com/news/366639658/Huge-grid-and-heat-challenges-ahead-as-Nvidia-set-for-1MW-rack">technical limitations of retrofitting</a> legacy commercial properties for high-density graphics processing unit (GPU) racks.</p> <section class="section main-article-chapter" data-menu-title="You recently gave evidence to the Environmental Audit Committee on the UK’s AI growth zone policy. What do you think is the realistic pipeline for datacentre capacity, given the government’s expectations?"> <h2 class="section-title"><i class="icon" data-icon="1"></i><span style="color: #34495e;">You recently gave evidence to the Environmental Audit Committee on the UK’s AI growth zone policy. What do you think is the realistic pipeline for datacentre capacity, given the government’s expectations?</span></h2> <p>The government’s expectation of multiple gigawatts is nonsense. With UK power costing roughly four times more than in the Nordics or the US, running a 100MW facility here would cost a quarter of a billion pounds more a year in power. Of the 119 applications currently in planning or under construction, we will realistically see only about two gigawatts reach the market by the middle of the next decade, not 50GW. The growth we get here will be purely organic or sovereign.</p> </section> <section class="section main-article-chapter" data-menu-title="How does the retail colocation business model differ from hyperscale development under these constraints?"> <h2 class="section-title"><i class="icon" data-icon="1"></i><span style="color: #34495e;">How does the retail colocation business model differ from hyperscale development under these constraints?</span></h2> <p>In retail colocation, we have to “build and they will come” to support the local ecosystem. You cannot wait for customers to line up. In contrast, the hyperscale market is different. These are billion-pound facilities; operators will not break ground on <a href="https://www.computerweekly.com/news/366642960/Data-dive-Power-grid-data-shows-birth-of-AI-in-UK-datacentres">a 100MW site</a> in the hope of leasing it – they wait until they have secured an anchor customer for a build-to-suit project.</p> </section> <section class="section main-article-chapter" data-menu-title="What is driving enterprises back to edge datacentres, and how does data sovereignty fit into this shift?"> <h2 class="section-title"><i class="icon" data-icon="1"></i><span style="color: #34495e;">What is driving enterprises back to edge datacentres, and how does data sovereignty fit into this shift?</span></h2> <p>A decade ago, there was a massive migration off-premises to larger datacentres and then to the cloud. Now we are seeing a movement back to the edge. Enterprises are <a href="https://www.computerweekly.com/feature/Is-cloud-data-sovereignty-all-just-a-case-of-Trust-me-bro">concerned about the US Cloud Act</a> and want to keep their data physically within their local jurisdiction. As a European-headquartered company, we provide that close-coupled, connected, sovereign local compute that global players cannot match.</p> <blockquote> <div class="imagecaption alignLeft"> <img src="https://cdn.ttgtmedia.com/rms/computerweekly/Dawn-Childs-nLighten-PR-140x180px.jpg" alt="Photo of nLighten CEO Dame Dawn Childs"> </div> <p><span style="font-size: 14pt;"><strong><span style="color: #34495e;">“Of the 119 [UK datacentre] applications currently in planning or under construction, we will realistically see only about two gigawatts reach the market by the middle of the next decade, not 50GW”</span></strong></span></p> <p><em><span style="color: #34495e;">Dawn Childs, nLighten</span></em></p> </blockquote> </section> <section class="section main-article-chapter" data-menu-title="If there is US software in your servers, or global equipment dependencies, doesn’t a backdoor risk remain regardless of where the facility is headquartered?"> <h2 class="section-title"><i class="icon" data-icon="1"></i><span style="color: #34495e;">If there is US software in your servers, or global equipment dependencies, doesn’t a backdoor risk remain regardless of where the facility is headquartered?</span></h2> <p>The concept of <a href="https://www.computerweekly.com/feature/Sovereign-cloud-and-AI-services-tipped-for-take-off-in-2026">sovereign AI</a> is ill-defined. You can never fully get away from the global supply chain. It is up to companies to define their sovereign requirements and do their own diligence. However, partnering with a company headquartered in Europe means workloads are not constantly shifted globally for efficiency, and that gives you a greater sense of control.</p> </section> <section class="section main-article-chapter" data-menu-title="You have upgraded the Bristol site to 1.2MW, but what happens when hardware demands 1MW per rack? Can legacy structures like this cope?"> <h2 class="section-title"><i class="icon" data-icon="1"></i><span style="color: #34495e;">You have upgraded the Bristol site to 1.2MW, but what happens when hardware demands 1MW per rack? Can legacy structures like this cope?</span></h2> <p>Some legacy facilities absolutely cannot be upgraded for those densities due to rigid ceiling heights or cooling routing constraints. However, we have fitted Bristol to scale up rack density using <a href="https://www.computerweekly.com/news/366644373/Data-dive-Dodgy-data-derails-datacentre-water-debate">direct-to-chip liquid cooling</a>. For the massive high-density compute demands of the future, we have modular, special-purpose facilities coming online designed to keep pace with emerging hardware.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about datacentres</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366643098/Bradford-datacentre-with-heat-reuse-gains-planning-consent">Bradford datacentre with heat reuse gains planning consent</a>: Deep Green’s 5.6MW AI datacentre will link up to an energy centre to heat buildings.</li> <li><a href="https://www.computerweekly.com/news/366640935/Data-dive-Government-2030-datacentre-capacity-targets-look-shaky">UK government’s 2030 datacentre capacity targets look shaky</a>: We look at UK datacentre capacity and find DSIT’s 6GW target is out of reach.</li> </ul> </div> </div> </section> We talk to the CEO of nLighten about the limits of the UK power grid, why that makes ‘edge’ datacentres a good idea, and navigating contemporary data sovereignty requirements https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/nLighten-Bristol-datacentre-CREDIT-Antony-Adshead-hero.jpg https://www.computerweekly.com/news/366645014/nLighten-CEO-Dawn-Childs-on-edge-datacentres-and-sovereignty Tue, 23 Jun 2026 05:40:00 GMT <p>Broadband network operator <a href="https://aureon.com/about/">Aureon</a>, Nokia and <a href="http://www.t3broadband.com/">t3 Broadband</a> have deployed an ultra-high-capacity artificial intelligence (AI)-optimised infrastructure optical transport network connecting a major datacentre development in North Dakota to the Chicago metro area.</p> <p>Headquartered in Kansas, t3 Broadband has the official stated mission to support service providers, utilities and communities of all sizes by helping them bring customers fast, reliable and secure broadband services.</p> <p>It claims it can “expertly” analyse, design and service economically sound broadband networks by using proven processes and technology.&nbsp;</p> <p>The company regards broadband as an economic driver for communities, adding that with the global economy evolving towards being always-connected, access to broadband connections is becoming an essential service.</p> <p>With Nokia’s optical networking service and delivered through t3 Broadband’s infrastructure expertise, the deployment is claimed to provide the advanced connectivity needed to support the rapid growth of AI, cloud and <a href="https://www.computerweekly.com/news/366622462/Nokia-networking-backbone-firms-up-ResetData-AI-factory">hyperscale</a> <a href="https://www.computerweekly.com/news/366622462/Nokia-networking-backbone-firms-up-ResetData-AI-factory">datacentre workloads</a>.</p> <p>The route is designed to support scaling up to 400Tbps of transport capacity, making it Aureon’s highest-capacity network connection to date and, according to Nokia, will create a scalable foundation for future AI infrastructure growth.</p> <p>Specifically, the route has at its heart <a href="https://www.nokia.com/optical-networks/1830-global-express/">Nokia’s 1830 Global Express platform</a>, Super C and L-band optical line system, and 1.2T ICE7 coherent optics.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about optical networking</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366641460/Optical-networks-to-bridge-the-AI-compute-consumption-gap">Optical networks to bridge the AI compute-consumption gap</a>: With AI spurring gigawatt-scale datacentre builds across APAC, Ciena is deploying ultra-fast, energy-efficient optical networking and AI-driven automation to ensure AI services can reach consumers.<i> </i></li> <li><a href="https://www.computerweekly.com/news/366635152/Microsoft-creates-framework-for-secure-optical-network-architecture">Microsoft creates framework for secure optical network architecture</a>: IT behemoth taps optical network provider to create ‘survivability by design’ architecture said to ensure operational continuity at all times, during both planned and unplanned outages.</li> <li><a href="https://www.computerweekly.com/news/366633494/Nokia-lands-railway-comms-optical-networking-wins-across-Asia">Nokia lands railway comms, optical networking wins across Asia</a>: Bangkok Expressway and Metro Public Company to deploy internet protocol/multi-protocol label switching technology, while GBI to build optical network for faster, more resilient connectivity across Middle East.</li> <li><a href="https://www.computerweekly.com/news/366631976/Brightsolid-expands-national-network-with-diverse-100Gbps-optical-connectivity">Brightsolid expands national network with diverse 100Gbps optical connectivity</a>: Strategic partnership issues assurance to deliver what it claims will be ‘ultra-fast’&nbsp;resilient connectivity from Scotland to London, supporting UK digital infrastructure growth.</li> </ul> </div> </div> <p>The installation is said to expand Aureon’s network capabilities, enabling the company to deliver transport services at a scale traditionally associated with hyperscale operators.</p> <p>“This deployment gives us the capacity and flexibility to support large-scale datacentre interconnect requirements while establishing a blueprint for future expansion,” said Aureon CEO George O’Neal. “Working with Nokia and t3 Broadband has enabled us to bring a new level of transport capability to our <a href="https://www.computerweekly.com/news/366625757/Cisco-Live-2025-The-network-critical-for-the-future-of-the-AI-era">network in the AI era</a>.”</p> <p>Chris Crowe, t3 Broadband CEO, added: “Our goal is to help customers build networks that are ready for what’s next. Together with Nokia, we worked closely with Aureon to deliver a high-capacity optical transport solution that supports its long-term growth strategy. This deployment strengthens Aureon’s ability to serve large-scale datacentre and AI connectivity requirements while providing a foundation for future expansion.”</p> <p>Matt Young, Nokia’s vice-president of sales, said: “AI is accelerating the need for scalable, high-performance connectivity between datacenters. Together with t3 Broadband and Aureon, Nokia is delivering the optical infrastructure needed to support that growth.</p> <p>“This deployment highlights how advanced optical networking can help operators efficiently scale capacity, maximise fibre resources and build networks ready for the demands of AI.”</p> <p>As it was working with Aureon to meet t3 Broadband’s needs, Nokia also announced that its defence division has forged a collaboration with pan-European <a href="https://knds.com/en">land defence company KNDS,</a> to deliver connectivity for soldiers and unmanned vehicles, addressing critical communication gaps on the modern battlefield.</p> <p>By integrating the Nokia Banshee Deployable Solution into the KNDS VBCI (Armored Infantry Fighting Vehicle), Nokia says it can enable seamless, real-time connectivity as forces transition from armoured vehicles into complex mission environments, supporting more responsive and coordinated operations.</p> Comms tech provider teams with US operator to provide AI-ready optical route connecting North Dakota datacentre to Chicago metro area https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/Nokia-PON-2024-PR-hero.jpg https://www.computerweekly.com/news/366644980/Nokia-t3-Broadband-Aureon-team-to-deploy-optical-AI-connectivity Tue, 23 Jun 2026 05:30:00 GMT <p>The escalation of conflict in the Middle East earlier this year <a href="https://www.computerweekly.com/resources/Data-centre-capacity-planning">disrupted datacentre operations</a> across <a href="https://www.computerweekly.com/news/366643123/How-geopolitical-instability-could-reshape-Gulf-datacentre-investments-and-sovereign-AI-strategies">parts of the Gulf</a>, with outages affecting financial services, government systems and mobile networks.</p> <p>For an industry built on the promise of continuous availability, it was a reminder of something fundamental. The risks operators plan for have widened – but the industry’s definition of resilience has not.</p> <h3>The resilience gap</h3> <p>Datacentre resilience is still largely defined in terms of uptime. Redundancy, backup generation and tier classifications remain essential. But they address a narrow set of conditions – primarily equipment failure and short-term power loss.</p> <p>That is no longer the full picture. Resilience today is the ability to anticipate disruption, absorb it, recover quickly and keep operating as conditions change. It is as much about how a facility behaves under pressure as it is about whether it fails at all.</p> <h3>Why resilience is getting harder</h3> <p>The pressures are compounding. Climate volatility is increasing the frequency and severity of extreme heat events, flooding and wildfire risk in regions where datacentres are concentrated.&nbsp;</p> <p><a href="https://www.computerweekly.com/news/366642960/Data-dive-Power-grid-data-shows-birth-of-AI-in-UK-datacentres">Grid constraints</a> are delaying new connections and making existing supply less predictable, particularly as AI-driven compute accelerates demand. Technological change like rising power densities, new cooling requirements, and shifting workload profiles also means that assumptions made about a facility at design stage may not hold for its intended operational life.</p> <p>Layered on top of this is geopolitical instability, which introduces supply chain disruption, energy market volatility and, as recent events have shown, direct physical risk to digital infrastructure.&nbsp;</p> <p>These are not isolated threats, they’re interconnected. For example, a climate event can strain a grid already weakened by geopolitical disruption. Operators who assess these risks in isolation are likely to underestimate their combined impact.</p> <h3>Designed upstream, not retrofitted later</h3> <p>One of the most consequential shifts the industry needs to make is moving resilience to the start of the decision-making process. Too often, it is treated as a compliance exercise or a post-design mitigation – something addressed after the site is selected, the architecture is fixed and the procurement is underway.&nbsp;</p> <p>This is the wrong sequence. Site selection is the most important resilience decision an operator will make. It determines exposure to climate hazards, grid reliability, <a href="https://www.computerweekly.com/news/366644373/Data-dive-Dodgy-data-derails-datacentre-water-debate">water availability</a>, the regulatory environment and interaction with neighbouring facilities. Once an operator commits to a site, the cost of compensating for poor resilience fundamentals rises sharply.&nbsp;</p> <p>The same applies to architecture and procurement. Designing for a fixed power density, a single cooling mode or a specific fuel source locks in assumptions that may not survive the next technology cycle. Resilience must be a foundational design input, not a retrofit.</p> <h3>Resilience in practice</h3> <p>What does this look like across the domains that matter most?</p> <ul class="default-list"> <li> <p>Site selection should incorporate multi-hazard screening – climate, seismic, flood, wildfire, grid reliability and water stress – as a standard input, not an optional due diligence step. Analysis should also assess whether a single event could disrupt multiple facilities simultaneously.</p> </li> <li> <p>Physical security requirements are expanding beyond perimeter fencing and access control. Facilities in some regions now need to account for threats that were previously considered unlikely, and security design should be integrated with operational continuity planning rather than treated as a standalone discipline.</p> </li> <li> <p>Power flexibility means moving beyond backup generators to diversified energy strategies – grid integration combined with on-site generation, battery storage, microgrids and load flexibility. The goal is not just backup, but the ability to operate flexibly within wider energy systems – including offering grid services such as frequency response and demand shifting.&nbsp;</p> </li> <li> <p>Water strategy must address operational and reputational risk. In water-stressed regions, reliance on high-consumption cooling exposes operators to regulatory intervention and community opposition. Practical responses include closed-loop systems, non-potable water sources, greywater reuse and, where appropriate, dry or hybrid cooling – with trade-offs assessed honestly against energy efficiency.</p> </li> <li> <p>Operational preparedness – people, processes, testing and governance – is the domain most often underinvested. Resilience is not only a design problem; it depends on trained teams, regularly tested response plans and governance structures that can make decisions under pressure.</p> </li> </ul> <h3>A practical resilience framework</h3> <p>Operators do not need to start from a blank page. A simple, four-stage cycle provides a practical starting point:</p> <ol class="default-list"> <li> <p>Know the risk. Model and assess multi–hazard screening, climate projections, grid reliability analysis, water stress mapping.</p> </li> <li> <p>Plan and design. Embed findings into site selection, architecture, procurement and operational planning.</p> </li> <li> <p>Respond. Develop and regularly test emergency and business continuity plans, including scenarios that combine multiple simultaneous stresses.</p> </li> <li> <p>Learn and adapt. Feed operational experience, incident data and changing external conditions back into the cycle. Resilience is not a fixed state. It is a continuous process.&nbsp;</p> </li> </ol> <p>This cycle applies equally to new builds and existing facilities, and it scales from individual sites to global portfolios.</p> <h3>The stranded-asset risk</h3> <p>One of the greatest threats to long-term value comes from inflexibility. Density, workloads and technologies are evolving faster than most datacentre assets can be rebuilt. Facilities designed around narrow assumptions risk becoming constrained or stranded well within their intended lifespan. Those assumptions might relate to fixed power density, a single cooling mode or a specific regulatory environment.</p> <p>A better approach is modular, adaptable design that includes standardised power and cooling blocks that can be reconfigured as requirements change; scalable architecture that accommodates rising densities without wholesale redesign, and; procurement strategies that favour flexibility over lowest initial cost.</p> <p>This is not speculative. The shift towards prefabricated, containerised and modular solutions is already accelerating across the sector, driven as much by delivery speed as by resilience. But the resilience dividend – the ability to adapt rather than replace – is significant and often undervalued in investment decisions.</p> <h3>What needs to change now</h3> <p>Operators need to broaden how they assess risk and treat geopolitical and climate pressures as part of the same picture. Resilience needs to move into early-stage decision-making, particularly around site selection and design assumptions.</p> <p>Power strategies must become more flexible. Water use needs to be actively managed, not passively assumed. Operational readiness needs to be tested against more than single-point failures.</p> <p>Most importantly, facilities need to be designed to adapt. Over the next decade, the datacentres that perform best will not be those designed simply to withstand disruption; they will be the ones designed to adjust to it.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about datacentre development</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366644373/Data-dive-Dodgy-data-derails-datacentre-water-debate">Data dive: Dodgy data derails datacentre water debate</a>. The Government Digital Sustainability Alliance reports that we are on track for a massive water supply shortfall. The dataset it used suggests not. We look at the datacentre water use debate</li> <li><a href="https://www.computerweekly.com/news/366643633/Datacentre-dive-From-rust-belt-to-megawatt-AI-factory">Datacentre dive: From rust belt to megawatt AI factory</a>. We visited Terawulf’s Lake Ontario 750MW datacentre development. Photos and recordings weren't allowed, so we took notes and wrote them up in more traditional ways</li> </ul> </div> </div> <p></p> <p></p> Datacentre resilience must evolve beyond uptime to anticipate interconnected climate, grid, and geopolitical risks by embedding adaptive design and strategy into early planning https://cdn.ttgtmedia.com/visuals/German/article/datacenter-monitoring-adobe.jpg https://www.computerweekly.com/opinion/Datacentre-resilience-means-more-than-uptime-Heres-what-to-change Tue, 23 Jun 2026 05:26:00 GMT <p>For years, the assumption around AI infrastructure was easy to accept. Serious compute would be built <a href="https://www.computerweekly.com/resources/Software-as-a-Service-SaaS">where hyperscale cloud</a>, developer density, and capital were already concentrated, namely California, Seattle, London, and a small circle of established technology hubs.</p> <p>There was a practical reason for that geography. Training and deploying AI at scale requires datacentres, compute, networking capacity, energy, and advanced infrastructure to work together.&nbsp;</p> <p>OECD analysis notes that this has pushed AI firms toward services operated by <a href="https://www.computerweekly.com/feature/Is-cloud-data-sovereignty-all-just-a-case-of-Trust-me-bro">the largest cloud computing suppliers</a>. Over time, that dependence hardened into <a href="https://www.computerweekly.com/feature/This-rise-of-the-splinternet-Data-sovereignty-risks-and-responses">market concentration</a>. In the third quarter of 2025, Synergy Research Group put Amazon, Microsoft, and Google's combined share of global enterprise cloud infrastructure spending at 63%.</p> <p>That logic now looks less durable. Compute is becoming more expensive, more power-intensive, and harder to access outside <a href="https://www.computerweekly.com/feature/Breaking-the-stranglehold-Responses-to-data-sovereignty-risk">a small group of dominant providers</a>. Builders are starting to confront questions that hyperscale cloud mostly let them ignore.&nbsp;</p> <p>Where will the power come from? Can chips be shipped to this jurisdiction? Whose laws apply to the data once it moves?&nbsp;</p> <p>Those questions are getting answered in different places now, and most of them are not in Silicon Valley.</p> <section class="section main-article-chapter" data-menu-title="What scarcity teaches"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What scarcity teaches</h2> <p>In established cloud markets, the default answer to rising AI demand is to add more capacity through larger cloud contracts, denser datacentre buildout, and deeper dependence on the same centralised stack.</p> <p>That answer is becoming harder to scale. Datacentres consumed about 1.5% of the world's electricity in 2024, which was enough to make energy one of the pressure points in AI infrastructure. The International Energy Agency expects that share to rise to just under 3% by 2030, making compute harder to treat as a hidden layer behind AI products.</p> <p>In much of the developing world, that pressure was already the starting point. Builders there have rarely had the option of treating compute access, power, and distribution as someone else's problem. They have had to design for it.&nbsp;</p> <p>The result is a quieter pattern that does not get much attention in Silicon Valley coverage. Namely that serious AI infrastructure is now being built in places where scarcity is treated as a design problem rather than an afterthought.</p> </section> <section class="section main-article-chapter" data-menu-title="What this looks like in practice"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What this looks like in practice</h2> <p>The pattern is most visible across four regions.</p> <p>In India, Yotta Data Services runs Shakti Cloud on more than 16,000 NvidiaH100 GPUs and is on track to roughly double that by the end of 2025. Over half of the compute behind the IndiaAI Mission – the government's push to build indigenous foundation models – sits on Yotta's hardware.&nbsp;</p> <p>In February 2026, the national multilingual platform Bhashini moved off foreign hyperscalers and onto Shakti Cloud, picking up roughly 40% in performance along the way. Bhashini runs real-time translation across 11 Indian languages at population scale and the people running it had decided that infrastructure they could not govern was the wrong place to put it.</p> <p>Across Africa, Cassava Technologies, founded by Zimbabwean entrepreneur Strive Masiyiwa, is deploying 12,000 Nvidia GPUs across datacentres in South Africa, Egypt, Kenya, Morocco, and Nigeria.&nbsp;</p> <p>Cassava is the first Nvidia Cloud Partner on the continent. Before this buildout, Nvidia estimated that roughly 80 of its GPUs were installed across the entire African continent. The constraint was not only compute pricing; it was the basic absence of advanced silicon.&nbsp;</p> <p>Cassava's response is a pan-African network running on its own fibre backbone, designed so that African startups, researchers, and governments do not have to route through Europe or the United States to train and deploy AI.</p> <p>In Brazil, the government's SoberanIA project reserves 500MW for a sovereign AI factory in Piauí, powered entirely by renewable energy, with Scala datacentres as lead infrastructure partner.&nbsp;</p> <p>Meanwhile, Brazil has committed to attracting up to $370 billion in datacentre investment over the next decade, tied to the REDATA program's tax incentives for projects sourcing 100% renewable power. Roughly 65% of Brazilian data is still stored abroad. The wager is that abundant hydroelectric and solar power gives Brazil the kind of compute the US and China have to work harder to build – clean by default, cheap by geography.</p> <p>The United Arab Emirates is taking the most expensive route. Core42, part of the G42 group, sells inference capacity on a mix of Nvidia and Qualcomm chips out of Abu Dhabi, and the country has committed jointly with the United States to a 10-square-mile, 5GW AI campus that should be partially operational by the end of the decade.&nbsp;</p> <p>The Emirati pitch is straightforward. Countries that want sovereign AI but cannot build the underlying stack themselves can rent one from a friendly government. The Middle East Institute describes it as a deliberate strategy of vertical integration – owning the chips, the power, the datacentres, and the foreign relationships in one piece.</p> <p>These projects do not share a politics or an ownership model. What they share is a starting assumption that compute access, power, land, and chip supply are first-order design problems rather than externalities. That assumption produces different infrastructure.</p> </section> <section class="section main-article-chapter" data-menu-title="Why inference changes the map"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Why inference changes the map</h2> <p>Training large models still rewards dense clusters, large capital budgets, and access to advanced chips. That work is unlikely to leave the largest hyperscale facilities soon.</p> <p>Inference is a different problem. Models are used continuously, by customers, devices, agents, and enterprise systems. McKinsey expects inference to overtake training in AI datacentres by 2030, to account for more than half of AI compute and roughly 30% to 40% of total datacentre demand.</p> <p>Inference asks different questions than training does. Rather than where the largest cluster can be built, the questions become where compute should sit, how fast it can respond, how reliably workloads can be routed, and whose laws govern the data while it does so. Those questions have geographic answers that hyperscale concentration does not handle well, especially for the billions of people who do not live within easy latency of a US or European datacentre.</p> <p>The compute fabric that inference demand requires is broader than hyperscale cloud alone can provide. Distributed GPU capacity, regional inference clusters, sovereign clouds, and emerging neoclouds in places such as Mumbai, Nairobi, São Paulo, and Abu Dhabi are not substitutes for hyperscale. They are the layer hyperscale cannot serve on its own.</p> </section> <section class="section main-article-chapter" data-menu-title="What this means for the map"> <h2 class="section-title"><i class="icon" data-icon="1"></i>What this means for the map</h2> <p>The old map of AI infrastructure was drawn around places where cloud capacity was already concentrated. That map made sense when compute was treated as cheap and abundant.</p> <p>The next map will look different. It will be drawn around places that learned to build when compute was costly and strategic, and where the question of who controls the stack was never theoretical. The companies and governments doing that work are not catching up with Silicon Valley. They arrived at the problem first, because they had to.</p> <p><em>Ilman Shazhaev is founder and CEO of Dizzaract, an AI infrastructure company headquartered in Abu Dhabi. He serves as a UN/UNODC expert panel member advising on AI applications in developing economies and has authored 46 scientific articles and 10 registered invention patents.</em></p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about datacentre pipeline</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/feature/Data-dive-A-new-American-Century-in-the-datacentre-pipeline">Data dive: A new American Century in the datacentre pipeline?</a> Looking at datacentre development internationally, we see how the UK faces apparent relative decline, how countries are responding to the AI age, and what MW vs GDP can tell us</li> <li><a href="https://www.computerweekly.com/news/366640447/Hit-the-north-UK-datacentre-focus-shifts-to-M62-and-points-north">Hit the north! UK datacentre focus shifts to M62 and points north</a>. Barbour ABI data shows 8GW of total datacentre pipeline with most big projects in the north and Scotland, while London and the M4 corridor are about 25% of projected capacity</li> </ul> </div> </div> <p></p> <p></p> </section> AI infrastructure is moving beyond hubs like Silicon Valley. Nations like India, Brazil, and the UAE are building sovereign, power-conscious capacity to solve local compute scarcity https://cdn.ttgtmedia.com/visuals/ComputerWeekly/Hero%20Images/Outsourcing-map-business-fotolia.jpg https://www.computerweekly.com/opinion/AIs-next-compute-layer-is-likely-to-come-from-outside-Silicon-Valley Mon, 22 Jun 2026 11:05:00 GMT <p>The <a href="https://www.computerweekly.com/news/366639995/Enormous-AI-growth-zone-datacentre-gets-planning-approval">rapid adoption of artificial intelligence (AI)</a> is colliding with a hard physical limit. Notably, that limit is not necessarily compute capacity or silicon availability, but electricity.</p> <p>For decades, technology leaders treated power as a basic utility. You plugged in your servers and the grid delivered. Yet in an AI-driven world, power is fast becoming the defining constraint on digital growth.&nbsp;</p> <p>Consider the fundamental difference between standard cloud workloads and generative AI.&nbsp;</p> <p>A standard internet search requires a fraction of the energy needed to process a single query through a large language model. Multiply that single interaction by billions of users worldwide and the true scale of the problem becomes clear. The computational density of AI workloads is creating highly concentrated energy demands that regional electricity grids were not built to handle.</p> <p>We are already witnessing the consequences of this shift. Gartner predicts that by 2026, power shortages will delay more than 30% of all <a href="https://www.computerweekly.com/resources/Data-centre-capacity-planning">datacentre expansions</a>. By 2028, datacentres will account for 10% of all US electricity demand. This is forcing a strategic reset <em>en masse</em> for IT and infrastructure &amp; operations (I&amp;O) leaders.</p> <section class="section main-article-chapter" data-menu-title="The emergence of the tech energy company"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The emergence of the tech energy company</h2> <p>The scale of the challenge is rather staggering. Some proposed hyperscale datacentres will consume up to 5GW of power – enough to power roughly 3.5 million households. When you concentrate that much demand into a few square miles, existing commercial grids can buckle. More recently, this is combining with increasing friction with local communities who fear grid instability and rising utility costs.</p> <p>Grid operators work on 20-plus year planning cycles, yet excitement for AI evolves and transforms month-to-month. Because <a href="https://www.computerweekly.com/news/366639772/National-Grid-Nebius-and-Emerald-hail-datacentre-power-throttling">the public grid cannot keep up</a>, major technology companies are taking matters into their own hands. Constellation and Microsoft have committed to restart the Three Mile Island nuclear plant by 2028, whilst Amazon has unveiled plans to invest $20 billion into a datacentre campus near a Pennsylvania power plant.</p> <p>These organisations are securing "behind the meter" power because they must. On-site power generation is transitioning from a redundancy to a primary source of power. Consequently, IT and I&amp;O leaders increasingly evaluate alternative energy sources to guarantee future capacity. Interest is growing in small modular reactors that promise safer and more flexible nuclear power deployment, whilst natural gas microgrids are being utilised to bridge the immediate gap while renewable technologies mature.</p> </section> <section class="section main-article-chapter" data-menu-title="Cooling the AI engine"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Cooling the AI engine</h2> <p>Generating power is only one part of the equation. Organisations must also radically optimise how they consume it. New construction facilities will house server racks that generate unprecedented levels of heat. Traditional air cooling physically fails at these new high-density thresholds and is no longer viable for high performance computing.</p> <p><a href="https://www.computerweekly.com/news/366644373/Data-dive-Dodgy-data-derails-datacentre-water-debate">Liquid cooling technologies are now the standard</a> for AI clusters, with systems that circulate liquid directly to (and away from) high heat components for better thermal performance. Implementing advanced techniques like rear door heat exchangers significantly reduces operational costs. It represents a fundamental shift in facility engineering.</p> <p>Ironically, while AI is driving the need for more cooling, it can also provide the solution. AI algorithms can analyse real time sensor data for workload and temperature to predict future cooling needs dynamically. This intelligent automation can result in up to a 40% reduction in cooling energy consumption.</p> </section> <section class="section main-article-chapter" data-menu-title="The role of sustainable software optimisation"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The role of sustainable software optimisation</h2> <p>The conversation around datacentre energy often ignores the software layer entirely. However, developers play a crucial role in managing this power crisis. Training large AI models is incredibly energy intensive. The industry must shift its focus toward writing more efficient code and right sizing models for specific tasks.</p> <p>By optimising algorithms to require fewer computational cycles, organisations can significantly reduce the underlying power draw of their applications. Choosing the most efficient software tool for the job is increasingly a vital energy decision.</p> </section> <section class="section main-article-chapter" data-menu-title="The edge computing advantage"> <h2 class="section-title"><i class="icon" data-icon="1"></i>The edge computing advantage</h2> <p>The industry has spent years centralising data processing to achieve economies of scale, yet power constraints are reversing that trend. We have reached a tipping point where the immense size of centralised facilities is becoming a liability.</p> <p>Edge computing offers a strategic alternative. IT leaders and I&amp;O leaders need to distinguish between training AI models and actually running them. While model training requires massive, centralised power clusters, day-to-day inferencing can often happen locally.&nbsp;</p> <p>By deploying more localised edge data centres, organisations can move data processing closer to the source. This decentralisation spreads the electrical demand across multiple local power grids instead of overwhelming a single region. These purpose-built facilities also require less energy to transmit data and can often utilise highly efficient free cooling techniques.</p> </section> <section class="section main-article-chapter" data-menu-title="Navigating the energy transition"> <h2 class="section-title"><i class="icon" data-icon="1"></i>Navigating the energy transition</h2> <p>The organisations that succeed in the AI economy will be those that recognise energy as a strategic variable rather than a facilities afterthought. Treating power availability as a given is a recipe for stalled growth. Those who fail to adapt their site selection and architecture design risk giving up competitive ground.</p> <p>The next phase of IT infrastructure will be defined by how well organisations can secure reliable power and intelligently distribute workloads. IT and I&amp;O leaders must bring their facility managers and software developers to the same table. Those that get this alignment right will lead the next generation of digital innovation.</p> <p><em>Gartner analysts will further explore datacentre power priorities and challenges at the Gartner IT Infrastructure, Operations &amp; Cloud Strategies Conference, taking place in London, from 16-17 November 2026.</em></p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about datacentres and electricity</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366643673/Datacentre-dive-AI-factory-power-draw-changes-the-grid-calculus">Datacentre dive: AI factory power draw changes the grid calculus</a>. We look at energy as the key driver – and bottleneck – in development, and why water use is less of an issue now datacentres use liquid cooling over air cooling</li> <li><a href="https://www.computerweekly.com/news/366642960/Data-dive-Power-grid-data-shows-birth-of-AI-in-UK-datacentres">Data dive: Power grid data shows birth of AI in UK datacentres</a>. Electricity supply utilisation ratios show datacentre developer ‘land grab’, capacity switch-ons, the coming of Hopper and Blackwell GPUs, and usage ramping during training runs</li> </ul> </div> </div> </section> AI growth is now hitting a hard limit – electricity. With power shortages causing delays, firms are pivoting to on-site energy, liquid cooling, and edge computing to sustain scaling for AI https://cdn.ttgtmedia.com/visuals/German/article/spark-blue-fotolia.jpg https://www.computerweekly.com/opinion/The-power-crunch-How-energy-constraints-reshape-datacentre-strategy Mon, 22 Jun 2026 10:02:00 GMT <p>On 1 March 2026, <a href="https://www.computerweekly.com/news/366639901/Iran-war-a-melting-pot-for-other-cyber-threats">Iran fired drones</a> into three Amazon datacentres in the UAE and Bahrain. Banks went dark. Payments stopped. Millions of people across <a href="https://www.computerweekly.com/news/366641798/Dubai-rolls-out-AI-training-for-50000-government-staff">Dubai</a> and Abu Dhabi could not hail a cab or check their balance. The cloud, it turns out, has a postcode. And a postcode can be bombed.</p> <p>These were not random targets. Iran's Islamic Revolutionary Guard Corps framed the strikes as a response to military and intelligence activity it believed was running through those facilities. And it was not entirely wrong. The line between commercial cloud infrastructure and active defence asset had already gone well before the first drone lifted off.&nbsp;</p> <p>Israel had been using <a href="https://www.computerweekly.com/news/366632809/Microsoft-admonished-for-role-in-facilitating-Gaza-genocide">Azure servers</a> in the Netherlands to store intercepted intelligence and run AI models. The Trump administration had been using AI systems to assist offensive planning. Once that is public knowledge, the datacentre becomes a legitimate target in the mind of any adversary. That is the world in which we now operate.</p> <p>The strikes exposed something the industry has been slow to say plainly. These are sprawling, visible buildings whose most critical components sit outside the server hall entirely. Knock out the chillers and the whole thing goes offline.&nbsp;</p> <p>The security model that hyperscale operators have built over the past two decades was designed for a specific threat. Namely, a person on the ground. Guards, fencing, cameras, access controls. None of that has any bearing on a kamikaze drone at altitude.</p> <p>The geography compounds the problem. Seventeen submarine cables run through the Red Sea. With the Strait of Hormuz closed and Houthi operations resuming, both of the region's main data chokepoints were in active conflict zones at the same time.&nbsp;</p> <p>That is not a coincidence of bad timing. It is a strategy. Disrupt the digital infrastructure and you disrupt the AI capability that modern military operations are starting to depend on. You also happen to take down the banking system, the ride-hailing apps and the payment terminals of an entire civilian population. The dual-use character of the cloud is a gift to anyone looking to cause maximum disruption at minimum cost.</p> <p>The industry's response so far has been to point at availability zones. The idea is that separating facilities within a region means a single failure does not take everything down. What the UAE and Bahrain attacks showed is that when multiple zones are hit simultaneously, the model collapses. The architecture was designed to survive one site failing, not a coordinated strike across a region. When the redundancy model and the threat model encounter each other in the real world, the redundancy model loses.</p> <p>There is another layer that gets less attention. Data localisation laws require organisations in many jurisdictions to keep their data within specific geographic boundaries. In practice that means disaster recovery backups often sit in the same region as the primary facility. If that region comes under attack, you cannot recover what was stored there. The legal framework designed to protect <a href="https://www.computerweekly.com/opinion/Can-data-sovereignty-become-a-liability-in-war">data sovereignty ends up trapping data in a conflict zone</a>.</p> <p>This is the point at which the architecture question becomes unavoidable. Concentration was always the trade-off the industry chose not to price. Thousands of servers under one roof is efficient in peacetime. It produces excellent economies of scale, competitive pricing and the kind of infrastructure metrics that look good in an investor deck. In conflict, it is a large, fixed, well-known address.</p> <p>The answer is not missile defence over datacentres, though some analysts have proposed exactly that. That treats the symptom. You are still protecting a single large target, and a sufficiently determined adversary will find its way past whatever perimeter you build around it. The answer is to stop building digital infrastructure like a cathedral and start building it like a field operation. Distributed, hardened, designed to lose a node without losing the network.</p> <p>Edge computing is not a new idea. The principle is straightforward – move compute closer to where data is generated and used, in smaller units, dispersed across a wider geography. What has changed is the urgency of the argument.&nbsp;</p> <p>The resilience case for edge infrastructure used to be made in terms of latency and data sovereignty. Both of those remain valid. But the March strikes added a third argument that is harder to ignore. Namely, that a distributed system cannot be neutralised in a single strike. There is no single building to target, no single power feed to cut.</p> <p>Modular edge units run on their own power, their own cooling, their own connectivity. They do not depend on the regional grid or a single fibre corridor. They can operate in isolation if the surrounding network goes down. They can be hardened to ballistic standards where the threat environment requires it. They can be deployed in weeks rather than the eighteen months it takes to commission a conventional data hall. And critically, losing one of them does not cascade into a regional outage. The architecture absorbs the blow.</p> <p>There is an irony worth noting. The military has understood this principle for a long time. No serious defence planner puts all of a theatre's command and communications capability in one building. The doctrine of dispersal, of redundancy by geography rather than by duplication within a single site, is basic field craft. The question is why civilian digital infrastructure was ever built on the opposite principle, and the answer is that peacetime economics rewarded concentration. That calculation has now changed.</p> <p>The policy window won’t stay open indefinitely. US and Asian manufacturers are not standing still. Government procurement, for defence, healthcare and financial services, should specify edge-first architectures now – before the next conflict proves the point again at greater cost.&nbsp;</p> <p>The Middle East didn’t create this problem. It just made it impossible to look away.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about datacentres and the middle east</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366643123/How-geopolitical-instability-could-reshape-Gulf-datacentre-investments-and-sovereign-AI-strategies">How geopolitical instability could reshape Gulf datacentre investments and sovereign AI strategies</a>. Rising tensions are forcing hyperscalers, governments and investors to reassess risk, resilience and infrastructure strategies as the Gulf positions itself as a global AI powerhouse</li> <li><a href="https://www.computerweekly.com/news/366639610/AI-workloads-force-a-fundamental-redesign-of-Middle-East-datacentres">AI workloads force a fundamental redesign of Middle East datacentres</a>. From hyperscale GPU clusters to sovereign AI ambitions, Huawei outlines how infrastructure must evolve to meet regional demand.</li> </ul> </div> </div> <p></p> <p></p> When the redundancy model and the threat model encounter each other in the real world, the redundancy model loses, says Adhum Carter Wolde-Lule, director at Prism Power Group https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/Dubai-skyline-day-UAE-IRStone-adobe.jpg https://www.computerweekly.com/opinion/Datacentres-are-a-great-target-and-AZs-dont-help-so-we-need-edge Mon, 22 Jun 2026 09:37:00 GMT <p>nLighten has completed a £15m refurbishment of <a href="https://www.computerweekly.com/resources/Data-centre-capacity-planning">its Bristol datacentre</a>, doubling its artificial intelligence (AI)-ready power capacity to 1.2MW to support regional enterprise workloads.</p> <p>The upgrade – which scales the site from 750kW and introduces a modular 2.4MW lithium-battery uninterruptible power supply (UPS) alongside <a href="https://www.computerweekly.com/news/366644373/Data-dive-Dodgy-data-derails-datacentre-water-debate">closed-loop cooling</a> – forms part of a wider UK expansion programme by nLighten, which is investing more than £100m to upgrade its national network.</p> <p>Located in Bristol’s established technology corridor at the Aztec West business park, the facility aims to support regional enterprises, such as in Bristol’s defence, aerospace and AI research clusters.</p> <p>The building was originally built in 1995 for Royal Sun Alliance and later by several other office occupiers. Conversion to a datacentre involved structural upgrades plus a mechanical and electrical infrastructure overhaul, the latest of which will see it expanded to 1.2MW. Current active IT load is approximately 180kW, so there’s lots of headroom for incoming customer deployments.</p> <p>The datacentre comprises 42 newly installed 47U racks with cold-aisle containment server cooling. The heat rejection system is a closed-loop chilled water design that consumes no water once it has been filled and routes externally to two Engie dry cooling units.&nbsp;</p> <p>According to nLighten engineers, these operate non-mechanically most of the time – ie, fan-only – but will add refrigeration during extreme heat, such as the current heatwave.&nbsp;</p> <p>To support the likely arrival of high-density AI compute, data hall plumbing is ready for direct-to-chip liquid cooling to be tapped in, which will allow the site to scale to high-density deployments of 150kW to 200kW per rack.</p> <p>In the event of mains failure, the UPS is designed to bridge the load for up to 20 minutes at 1.2MW, while standby diesel generators are configured to start and take the full load on the bus bar within 18 seconds.&nbsp;</p> <p>Operational resilience is maintained through weekly engineering drills and quarterly mains-fail transfer tests, to ensure the engineering team is prepared for emergency procedures.</p> <p>nLighten is positioning the Bristol facility as a “sovereign infrastructure hub” in an effort to target public sector organisations and private enterprises with <a href="https://www.techtarget.com/searchcloudcomputing/definition/data-residency">data residency</a>, security and long-term operational resilience requirements.</p> <p>Explaining the focus on sovereignty and legacy upgrades, Dawn Childs, CEO of nLighten, said: “This is part of our mission to improve sustainability, to improve connectivity, to improve the network of edge datacentres. Regarding sovereignty, we are in the vanguard of that because we are taking legacy datacentres and enabling edge deployments so you can really know where your data is, what’s being done with it.”</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about datacentres</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366643098/Bradford-datacentre-with-heat-reuse-gains-planning-consent">Bradford datacentre with heat reuse gains planning consent</a>: Deep Green’s 5.6MW AI datacentre will take 24 months to build and will link up to an energy centre to heat buildings across Bradford city centre via pre-laid pipes.</li> <li><a href="https://www.computerweekly.com/news/366640935/Data-dive-Government-2030-datacentre-capacity-targets-look-shaky">Data dive – UK government’s 2030 datacentre capacity targets look shaky</a>: We look at UK datacentre capacity – current and projected – and find DSIT’s 2030 target for 6GW of AI-capable capacity is currently out of reach, unless operators get a move on.</li> </ul> </div> </div> <p>While edge datacentres – smaller facilities, often in repurposed buildings – can provide customers with local access and some <a href="https://www.computerweekly.com/feature/Is-cloud-data-sovereignty-all-just-a-case-of-Trust-me-bro">peace of mind in the sovereignty department</a>, converting older commercial infrastructure into high-density AI environments is likely to present physical limitations.&nbsp;</p> <p>Although nLighten’s Bristol facility has integrated cold-aisle containment and direct-to-chip liquid cooling readiness, legacy structures often suffer from rigid ceiling heights and floor load constraints. As Nvidia’s roadmap introduces rack <a href="https://www.computerweekly.com/news/366639658/Huge-grid-and-heat-challenges-ahead-as-Nvidia-set-for-1MW-rack">designs that will demand up to 1MW of power draw</a>, and therefore heat creation, retrofitting older spaces becomes exponentially difficult while designing adequate heat rejection pathways within confined footprints will be an engineering hurdle.</p> <p>Furthermore, regional expansions are heavily dependent on local grid availability. While a 5MVA grid connection provides sufficient near-term capacity for nLighten’s current 1.2MW design, securing additional power allocation in grid-constrained UK regions is a complex process.&nbsp;</p> <p>nLighten is backed by global infrastructure investor I Squared Capital and manages a European footprint of 36 carrier-neutral facilities across eight markets, totalling 67MW of capacity. This expansion has been accelerated by strategic acquisitions, including French platform Euclyde Data Centers in June 2023 and Proximity Data Centers in the UK in September 2023.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">The strategic value of edge datacentres</h3> <p>nLighten CEO Dawn Childs believes the industry is witnessing a migration of workloads from centralised public cloud architectures back to regional edge environments. This transition is heavily driven by digital sovereignty, with organisations increasingly concerned about jurisdictional exposure, such as the US Cloud Act and FISA Section 702 warrantless data requests.</p> <p>Childs said that while cloud platforms offer generic compute, they often lack the close-coupled, local control that sovereign European operators can provide.</p> <p>“What you do have control over is who you choose to partner with,” said Childs, suggesting that selecting European-headquartered providers helps enterprises avoid the <a href="https://www.computerweekly.com/news/366642487/Cloud-and-data-sovereignty-caught-in-a-paradox">jurisdictional entanglement of US-owned cloud platforms</a>, where data is routinely moved globally to maximise operational efficiency.</p> <p>The economic reality of the UK power market also shapes this edge strategy. Childs points out that UK grid electricity costs are approximately four times higher than in the Nordics or the US. For a massive 100MW AI facility, this differential adds roughly £250m a year in operating expenditure just for power. Consequently, Childs believes the UK is an unattractive market for GW-scale AI factories, and although the datacentre pipeline is in the region of 6GW to 8GW, only about 2GW will actually reach the market. The growth will instead be driven by organic, regional deployments, she believes.</p> <p>To support this local model, nLighten is focusing on community integration. The company has the capability for its UPS systems to absorb and release power to balance the local grid during sudden surges. The Bristol facility is also exploring partnerships with local heat networks. “These things take time. They don’t happen overnight,” said Childs.&nbsp;</p> </div> </div> European datacentre operator doubles potential AI-ready power capacity to 1.2MW with dry cooling at Bristol site as part of a wider £100m-plus UK modernisation programme https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/Bristol-NXiao-getty-RF.jpg https://www.computerweekly.com/news/366644836/nLighten-completes-15m-refurbishment-of-Bristol-edge-datacentre Fri, 19 Jun 2026 10:04:00 GMT <p>Google Cloud went big on its shift to agentic in its artificial intelligence (AI) strategy, where it set out its vision to move to agentic enterprise automation at its London Summit this week. That transition comes with a raft of tools – mostly announced earlier this year at <a href="https://www.computerweekly.com/opinion/What-did-we-learn-at-Google-Cloud-Next-2026">Google Cloud Next</a> – designed to allow organisations to move to multi-agent systems that can reason and execute complex business workflows autonomously.</p> <p>Speaking at the event at Tobacco Docks, Maureen Costello, vice-president for the UK, Ireland, and Sub-Saharan Africa at Google Cloud (<em>pictured above</em>), told attendees the era of experimentation is being supplanted by a drive for operational deployment.&nbsp;</p> <p>According to Costello, AI agents got but one mention at the previous year’s event, but have now become the central pillar of the provider’s roadmap as it seeks to scale multi-agent enterprise systems.</p> <p>“The agentic enterprise is happening right here, right now,” she said. “Leading UK businesses are moving past basic pilots. They’re deploying multi-agent systems – agents that can think, reason, collaborate and execute complex business workflows.”</p> <p>The <a href="https://www.computerweekly.com/resources/Software-as-a-Service-SaaS">slew of tools</a> fell into three core product pillars: frontier models, agent platforms and personal productivity tools.&nbsp;</p> <p>Firstly, there is a refresh of the Gemini frontier model family. Google confirmed that Gemini 3.5 Flash – a model performance-optimised for agents and coding tasks – will be available to UK customers by the end of June. This will be followed by Gemini 3.5 Pro later in the year and Gemini Omni, a multi-modal model designed for extreme speed and low latency across text, audio and video inputs.</p> <p>The provider also introduced a series of agentic platforms aimed at developers and business users. The Gemini Enterprise Agent Platform is targeted as a centralised mission control for managing autonomous workflows. It is complemented by Agent Designer v2, a low-code tool that uses natural language to build agents across disparate applications. For engineering teams, Antigravity 2.0 brings an agent-first code environment available via desktop or command-line interface.</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> The agentic enterprise is happening right here, right now,. Leading UK businesses are moving past basic pilots. They’re deploying multi-agent systems – agents that can think, reason, collaborate and execute complex business workflows </figure> <figcaption> <strong>Maureen Costello, Google Cloud</strong> </figcaption> <i class="icon" data-icon="z"></i> </div> </blockquote> <p>The expansion into the personal productivity space was marked by Gemini Spark, a 24/7 personal assistant intended for enterprise laptops and mobile devices. These tools are underpinned by the Knowledge Catalogue, which Google describes as a universal context engine designed to index and categorise enterprise data to make it accessible for AI reasoning.</p> <p>Meanwhile, on the hardware and security front, the provider unveiled its eighth-generation Tensor Processing Units (TPUs), custom AI hardware designed to improve efficiency for training and inference. This was paired with the launch of Google AI Threat Defence, a cyber security solution that uses autonomous agents to monitor for and mitigate AI-powered threats in real time.</p> <section class="section main-article-chapter" data-menu-title="How Unilever is using the Gemini Enterprise Agent Platform"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How Unilever is using the Gemini Enterprise Agent Platform</h2> <p>During the keynote, Google showcased its strategic partnership with Unilever, <a href="https://www.computerweekly.com/news/366639143/Google-Cloud-supplants-Azure-as-Unilever-cloud-of-choice">signed earlier this year</a>. The consumer goods giant is using the Gemini Enterprise Agent Platform to drive brand discovery and marketing. The deal is notable as it marks a significant expansion of Google’s footprint within Unilever, and effectively supplanted legacy Microsoft-centric ecosystems for advanced AI workloads.</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> We’re working with Google to go deeper on agentic commerce, building our guardrails and our observability layer as we embed AI across our global business </figure> <figcaption> <strong>Sam Kini, Unilever</strong> </figcaption> <i class="icon" data-icon="z"></i> </div> </blockquote> <p>Sam Kini, chief digital and technology officer at Unilever, explained that the company is systematically unpacking its core business processes through an agentic lens. Kini noted that the path to purchase is fundamentally changing, and requires organisations to become “AI native” to stay relevant in emerging commerce channels.</p> <p>“We believe that having the right balance of pace and control is key,” said Kini. “You have to be brave and you have to be prepared to throw out the playbook. We’re working with Google to go deeper on agentic commerce, building our guardrails and our observability layer as we embed AI across our global business.”</p> </section> <section class="section main-article-chapter" data-menu-title="How Google Cloud is investing in the British economy"> <h2 class="section-title"><i class="icon" data-icon="1"></i>How Google Cloud is investing in the British economy</h2> <p>The provider also showcased its ongoing commitment to UK infrastructure, centred on the scheduled opening of <a href="https://www.computerweekly.com/news/366631306/Waltham-Cross-to-get-heating-from-Google-datacentre">a datacentre in Waltham Cross, Hertfordshire</a>, later this year. This project forms part of a broader £5bn investment programme in the UK’s AI economy, which also includes a partnership to upskill 100,000 civil servants in AI technologies by 2030.</p> <p>“Sovereignty isn’t just a policy requirement. It requires physical steel in the ground. Google Cloud is investing heavily in the British economy ... ensuring that every British organisation has the tools, talent and technology to relentlessly innovate,” said Costello.</p> <blockquote class="main-article-pullquote"> <div class="main-article-pullquote-inner"> <figure> The ‘wait and see’ era of AI is over. If there was one word I could bring to mind that really looks at how the UK is behaving right now, it is acceleration. We are shifting from writing code to having agents build and deploy code for us at scale </figure> <figcaption> <strong>Alex Rutter, Google Cloud EMEA</strong> </figcaption> <i class="icon" data-icon="z"></i> </div> </blockquote> <p>To assist with the governance of these sprawling agent environments, Google introduced Agent Gateway. This tool serves as a single control point to govern first- and third-party agents across multiple clouds, providing OTEL (Open Telemetry)-compliant telemetry and granular “agent identity” tracking. This allows administrators to treat agents like a digital workforce, to monitor dependencies and actions.</p> <p>Alex Rutter, managing director for AI at Google Cloud EMEA, argued that the “wait and see” era of AI is over for the UK market, adding that the focus is now on “AI fluency” and the ability of organisations to deploy industry-specific agents that solve problems at the edge.</p> <p>“The ‘wait and see’ era of AI is absolutely over,” said Rutter. “And if there was one word I could bring to mind that really looks at how the UK is behaving right now, it is acceleration. We are shifting from writing code to having agents build and deploy code for us at scale.”</p> <p>Rutter further noted that the addition of Agent Observability provides deterministic instrumentation across the entire agent flow, giving businesses the absolute visibility they need into why an agent is behaving a certain way before it impacts production.</p> <p>As UK businesses transition from proofs of concept to production-ready agentic solutions, the competition between hyperscalers is intensifying. By positioning its “unified stack” – from custom TPUs to high-level agent platforms – Google is attempting to capture the next wave of enterprise spending by offering a more integrated, “agent-first” alternative to traditional productivity suites.</p> <div class="extra-info"> <div class="extra-info-inner"> <h3 class="splash-heading">Read more about Google Cloud and AI agents</h3> <ul class="default-list"> <li><a href="https://www.computerweekly.com/news/366644382/Ineffable-Intelligence-strikes-Google-Cloud-deal-for-Vera-Rubin-GPU-power">Ineffable Intelligence strikes Google Cloud deal for Vera Rubin GPU power</a>: London-based ‘anti-LLM’ developer led by AlphaGo founder David Silver selects Google Cloud Vera Rubin GPU infrastructure to build reinforcement learning ‘superlearners’.</li> <li><a href="https://www.computerweekly.com/news/366644633/Google-Cloud-Summit-UK-to-deploy-AI-powered-planning-system">Google Cloud Summit: UK to deploy AI-powered planning system</a>: The UK planning system is being reworked with artificial intelligence and computer vision to provide data in a consistent format.</li> </ul> </div> </div> </section> Hyperscaler prioritises process automation in UK showcase, with frontier models, agent platforms and development tools to the fore, with customers such as Unilever in the spotlight https://cdn.ttgtmedia.com/visuals/ComputerWeekly/HeroImages/Maureen-Costello-CREDIT-Antony-Adshead-hero.jpg https://www.computerweekly.com/news/366644938/Google-Cloud-boosts-for-enterprise-agentic-at-London-Summit Fri, 19 Jun 2026 05:34:00 GMT 60 editor@computerweekly.com