I hate it when I’m right; back in 2023 I wrote:

(3) If the Government is encouraged to suppress general access to code, it will attempt to do so; and then the suppressions will fail several years, perhaps a few decades, later.

…comparing this to the old ITAR regime; see also my more recent blogpost on the Regulation of Public [AI] Power; but anyway: via the LinkedIn of Lukasz Olejnik, we learn about this new draft bill:

This Act may be cited as the ‘‘Decoupling America’s Artificial Intelligence Capabilities from China Act of 2025’’.

So, DAAICCA? — wherein, from the mind of Josh Hawley, we find:

PROHIBITIONS ON IMPORT AND EXPORT OF ARTIFICIAL INTELLIGENCE OR GENERATIVE ARTIFICIAL INTELLIGENCE TECHNOLOGY OR INTELLECTUAL PROPERTY

…and restrictions upon:

…RESEARCH AND DEVELOPMENT THAT BENEFITS THE PEOPLE’S REPUBLIC OF CHINA…

…and:

PROHIBITION ON UNITED STATES PERSONS HOLDING AN INTEREST IN OR PROVIDING FINANCING TO CHINESE ENTITIES INVOLVED IN ARTIFICIAL INTELLIGENCE RESEARCH AND DEVELOPMENT.

So basically yeah, some Americans are proposing a huge export-control-based global censorship regime like ITAR all over again, and it will fail the same way if not beforehand / if not also causing more harm to America than good.

Lukasz’ Summary

U.S. Congress will consider a bill proposing 20 years in jail or $1M fines for individuals and $100M for companies downloading Chinese AI models. It would also ban working with or investing in Chinese-controlled AI entities. https://t.co/4QZ1WEGBft pic.twitter.com/h5nQ7FlGRA

— Lukasz Olejnik (@lukOlejnik) February 3, 2025