US Senators Demand Answers From X, Meta and Alphabet About Sexual Deepfakes

The problem of sexualized deepfakes in the tech world is now bigger than just X.

In a letter to the leaders of

The senators also required the companies to preserve all documentation and information related to the creation, discovery, supervision, and monetization of AI-generated sexual images, as well as any related policies.

The message comes hours after (X and xAI are part of the same company.)

Pointing to media reports about how easily Grok produced sexual and nude images of women and children, the senators suggested that platforms’ guardrails to prevent users from posting non-consensual sexual images may not be enough.

“We recognize that many companies have policies against non-consensual intimate images and sexual exploitation, and that many AI systems claim to block explicit pornography. But in practice, as we saw in the examples above, users find ways around these guardrails. Or these barriers fail,” the letter said.

Grok, and by extension X, has come under fire for enabling this trend, but other platforms are not immune.

Deepfakes first gained popularity on Reddit, when a page featuring synthetic porn videos of celebrities went viral before the platform took them down in 2018. Sexual deepfakes targeting celebrities and politicians have multiplied on TikTok and YouTube, though they usually originate elsewhere.

Meta’s oversight board last year subpoenaed two instances of explicit AI images of female public figures, and the platform has also allowed nudity apps to sell ads on its services, though it later sued a company called CrushAI. There have been multiple reports of children posting fake photos of their peers on Snapchat. Telegram, which was not included in the senators’ list, has also become notorious for hosting bots designed to remove photos of women.

X, Alphabet, Reddit, Snap, TikTok and Meta did not immediately respond to requests for comment.

The letter asks companies to provide:

• Policy definitions of “deep fake,” “non-consensual intimate images,” or similar terms.
• Descriptions of companies’ policies and enforcement approaches regarding non-consensual AI-assisted deepfakes of people’s bodies, non-nude images, altered clothing, and “virtual nudity.”
• A description of current content policies addressing edited media and explicit content, as well as internal guidance provided to moderators.
• How current policies govern AI tools and image generators as they relate to suggestive or intimate content.
• What filters, guardrails or measures have been implemented to prevent the generation and distribution of deepfakes?
• What mechanisms do companies use to identify deepfake content and prevent it from being re-uploaded?
• How do they prevent users from benefiting from this content.
• How platforms prevent themselves from monetizing non-consensual AI-generated content.
• How companies’ terms of service enable them to block or suspend users who post deepfake content.
• What are companies doing to notify victims of non-consensual sexual deepfakes?

The letter was signed by Senators Lisa Blunt Rochester (D-Delaware), Tammy Baldwin (D-Wis.), Richard Blumenthal (D-Connecticut), Kirsten Gillibrand (D-N.Y.), Mark Kelly (D-Ariz.), Ben Ray Lujan (D-Wis.), Brian Schatz (D-Hawaii), and Adam Schiff (D-Calif.).

The move comes just one day after xAI owner Elon Musk said he was “not aware of any nude images of minors created by Grok.” Later on Wednesday, California’s attorney general opened an investigation into xAI’s chatbot, after mounting pressure from governments around the world angry that there were no guardrails around Grok that allowed it to happen.

xAI confirmed that it is taking action to remove “illegal content on X, including… [CSAM] and non-consensual nudity,” although neither the company nor Musk addressed the fact that Grok was allowed to make such edits in the first place.

The problem is also not limited to sexual images that have been manipulated without their consent. Although AI-based image creation and editing services do not allow users to “undress,” they do allow one to easily create deep fakes. To pick a few examples, OpenAI’s Sora 2 reportedly allowed users to create explicit videos featuring children; Google’s Nano Banana appears to have created an image showing Charlie Kirk being shot; The racist videos created using Google’s AI video model have garnered millions of views on social media.

The problem gets even more complicated when Chinese photo and video generators come into the picture. Many Chinese technology companies and apps — especially those linked to ByteDance — offer easy ways to edit faces, voices, and videos, and these outputs have spread to Western social media platforms. China has stronger requirements for labeling synthetic content that the United States does not have at the federal level, with audiences relying instead on fragmented and questionably enforced policies from the platforms themselves.

US lawmakers have already passed some legislation seeking to rein in fake pornography, but the impact has been limited. The Take It Down Act, which became federal law in May, aims to criminalize the creation and dissemination of non-consensual sexual images. But a number of provisions in the law make it difficult to hold image-generating platforms accountable, because they focus most scrutiny on individual users instead.

Meanwhile, a number of states are trying to take matters into their own hands to protect consumers and elections. This week, New York Governor Kathy Hochul proposed laws that would require AI-generated content to be labeled as such, and ban non-consensual deepfakes in specific periods leading up to elections, including depictions of opposition candidates.