Sextortion & Artificial Intelligence
June 17, 2024
Sextortion is present in the lives of adolescents.
When you saw the images of Katy Perry at the MET Gala, did you think they were real? They weren’t. While this may seem a comical use of artificial intelligence (AI), the fact remains that someone created fake images of Katy Perry realistic enough to fool a large audience. Now, let’s imagine how this seemingly innocent usage of AI can be turned into something harmful.
Imagine that instead of AI being used to create Katy Perry, it is used to create photos of a minor at the local high school. Instead of a fancy MET Gala image, it is used to create an explicit photo. Once the photo is generated, a predator can send it the minor and threaten to send it to everyone in their high school, unless they send money. All of a sudden the innocent, playful use of AI has taken a dark turn into sextortion. If this use of AI has never crossed your mind, that’s okay. This blog will delve into the concerns with AI, and its impact on sextortion.
THINGS TO CONSIDER
New technology continues to evolve and perpetrators and criminals have found ways to utilize it. AI has given predators the resources to alter innocent images found online into seemingly real sexually-explicit photos. These unsavory people can now create explicit photos, rather than depending on the victim to send them the photos themself. To make matters worse, the technology used by the perpetrators has become increasingly easier to assess and use1.
The FBI urges the public to exercise caution when posting or direct-messaging personal photos, videos and other identifying information on social media, dating apps and other online sites2. Perpetrators can use artificial intelligence to create a deepfake, an image or recording that has been convincingly altered and manipulated to misrepresent someone as doing or saying something that was never actually done or said3. Moreover, these deepfake images can depict an individual in a situation they were never truly in. So, perpetrators can utilize AI to generate deepfakes of potential victims for the purpose of sextortion. Once a deepfake is created, the perpetrator can reach out to the victim and threaten to expose the fake—but realistic—images or videos.
Voice-cloning is another form of deepfake than can be created with AI, whereby software allows individuals to sound like another person. As a result, fake voices can be used to deceive people by impersonating loved ones4.The ability to voice-clone is becoming just as accessible as creating false images. A simple Google search shows that one can access voice-cloning services for as little as $55. Although voice-cloning has been reported involving scams such as kidnappings, it can be used for sextortion, as well. For example, a much older predator could use this software to “talk” to a 14-year-old girl, while sounding every bit like a 14-year-old boy.
LEGISLATION
Many states have passed legislation aimed at the use of deepfakes. In 2024 alone, 20 bills have been enacted in Alabama, Florida, Idaho, Indiana, Iowa, Kentucky, Mississippi, New Mexico, Oklahoma, Oregon, South Dakota, Tennessee, Utah and Washington6. Although some of these bills highlight deepfakes and AI related to political advertisement, it is also crucial to consider how these bills could be a steppingstone for passing future bills related to AI’s use in sextortion.
Utah — Criminal Offense Amendments: Amends the definition of counterfeit intimate image to include generated depictions6.
Washington — Fabricated Intimate or Sexually Explicit Images: Concerns fabricated intimate or sexually explicit images and depictions6.
New Mexico — Campaign Reporting Act: Provides that if a person creates, produces or purchases an advertisement that contains materially deceptive media, which includes, but is not limited to AI, the advertisement shall include a disclaimer; creates the crime of distributing or entering into an agreement with another person to distribute materially6.
HOW WE CAN HELP
Truth Intelligence Consulting training and curriculum development incorporates scientific principles of social dynamics, cognition and memory, well-being and communication into the interview process. Compared to traditional routine questioning methods, our questioning paradigm helps obtain truthful, accurate and detailed information from interviewees.
Aleczia Gingold
John Jay College of Criminal Justice, Forensic Psychology Department
John Gervino
CEO Truth Intelligence Global Consulting
REFERENCES
1 LMNTRIX. (2023, July 22). Sextortion and Artificial Intelligence – LMNTRIX. https://lmntrix.com/blog/sextortion-and-artificial-intelligence/
2Malicious actors manipulating photos and videos to create explicit content and sextortion schemes. (n.d.). https://www.ic3.gov/Media/Y2023/PSA230605
3deepfake. (2024). In Merriam-Webster Dictionary. https://www.merriam-webster.com/dictionary/deepfake#:~:text=deep%C2%B7%E2%80%8Bfake%20%CB%88d%C4%93p%2D%CB%8Cf%C4%81k,not%20actually%20done%20or%20said
4Voice cloning fraud & Cyber law. (n.d.). https://www.asianlaws.org/blog-post.php?url=voice-cloning-fraud-and-cyber-law
\5Saeidi, M. (2024, May 17). Voice cloning scams are a growing threat. Here’s how you can protect yourself. CBS News. https://www.cbsnews.com/newyork/news/ai-voice-clone-scam/