So, the bad news. Last week, Nonconsensual, sexually explicit deepfakes of Taylor Swift went viral on X, amassing millions of views and likes before the account was suspended. The images were created using AI technology and originated from a website known for publishing fake nude images of celebrities.
Despite platform policies against such content, many of the offending posts remain live, and new graphic fakes continue to appear. One of the most viral explicit posts, shared by a user with a blue check depicting Swift, received more than 45 million views and 24,000 reposts before removal. The images may have originated in a Telegram group, and X is being criticized for not taking swift action to remove them. Swift’s fans launched a mass-reporting campaign to have the images taken down.
X has confirmed that it blocked Taylor Swift searches on its platform to prioritize safety after pornographic deepfakes of the singer went viral. The move has drawn criticism for being slow to curb the spread of explicit images. X stated that it has a zero-tolerance policy towards non-consensual nudity and is actively removing the images and taking action against the responsible accounts. The origin of the images was traced to a Telegram group known for creating nonconsensual AI-generated images. Microsoft CEO Satya Nadella acknowledged the issue and emphasized the need for guardrails around technology to ensure safer content production.
No federal law makes it illegal to create deepfakes nationally, though some lawmakers have introduced bills to try.
Business Insider has a number of the laws being examined, where 14 states have introduced legislation around AI and deepfakes with elections. The White House is calling for legislation to protect people from fake sexual images generated by AI. White House press secretary Karine Jean-Pierre emphasized the need for legislative action and highlighted the role of social media platforms in enforcing their own rules to prevent the spread of such material.
The SAG-AFTRA union condemned the images as well. The union agreed to some vague AI consent terms.
And if you want a deep dive into it, Casey Newton has an article entitled “The Taylor Swift deepfakes are a warning” on Platformer.
Why do we care?
If you need a frame of reference to discuss with your customers, here’s an easy one. It contains just about everything you need. It’s a clear example of misuse, it’s an example that common sense says should be illegal yet is not, it shows ethical boundaries that need to be determined for organizations, and it also shows the legal peril that might come. As highlighted by Microsoft CEO Satya Nadella, there is a pressing need for technology guardrails to ensure safer content production. IT service providers should consider integrating safety measures into their AI offerings, like watermarking synthetic media or developing detection algorithms for deepfakes, or pressuring their vendors to provide them.
The lawsuits Taylor Swift can enact may shape AI policy pretty quickly. Let’s get into some of that next….