A recent letter submitted to the Federal Trade Commission (FTC) and attorneys general from all 50 states, alongside the District of Columbia, urges a thorough examination of Elon Musk’s Grok platform, focusing particularly on its new “Imagine” feature designed for AI-generated images and videos.
Launched earlier this month by xAI, the tool includes a “Spicy” mode that incentivizes users to produce not-safe-for-work (NSFW) content. During an initial test by Technology News, the feature unexpectedly generated topless deepfake videos of Taylor Swift without any prompting.
The letter was initiated by the Consumer Federation of America (CFA) and supported by 14 other consumer advocacy groups, such as the Tech Oversight Project, the Center for Economic Justice, and the Electronic Privacy Information Center (EPIC). It references Technology News‘s findings about the inappropriate deepfakes created by Grok, noting that while the platform currently does not allow user-uploaded photos in Spicy mode — which raises serious concerns about issues like revenge porn — it still generates nude videos using images from the tool itself. This capability can produce videos that resemble real individuals, posing risks for those depicted, particularly minors.
The correspondence further warns that if the current restrictions on user-uploaded photos are lifted, it would potentially ignite a wave of non-consensual deepfakes. The organizations caution that the platform’s management shows a tendency to eliminate moderation features under the pretext of promoting “free speech.” Although AI-generated nude images are illegal to be distributed knowingly under the Take It Down Act in the United States, those regulations may not apply directly to Grok. The signatories called on the FTC and other authorities to investigate xAI for possible breaches of laws regarding Non-Consensual Intimate Imagery.
Concerns were also voiced over the platform’s mechanisms preventing minors from accessing the Spicy mode, which only involves a pop-up asking users to confirm their age. Additionally, it was pointed out that one of the age selection pop-ups defaults to “2000,” raising concerns about compliance with the Children’s Online Privacy Protection Act and various state-specific regulations pertaining to age verification for adult content.
