US senators demand answers from tech giants on deepfake protections

A group of US senators is pressuring tech companies—including X, Meta, Alphabet, Snap, Reddit, and TikTok—to provide evidence of their policies and enforcement regarding the rise of sexualized deepfakes on their platforms. In a letter to the leaders of these companies, the lawmakers asked for proof of robust protections and policies to curb the spread of non-consensual intimate imagery.

The senators demanded that the companies preserve all documents and information relating to the creation, detection, moderation, and monetization of sexualized, AI-generated images. Specifically, the letter requests documentation on:

• Policy definitions of deepfake content and non-consensual intimate imagery.
• Enforcement approaches for nonconsensual AI deepfakes, including altered clothing and virtual undressing.
• How current content policies and filters prevent the generation and distribution of suggestive or intimate content.
• Mechanisms used to identify deepfakes, prevent re-uploads, and stop both users and platforms from monetizing this content.
• Terms of service enforcement and victim notification procedures.

The push comes as platforms face scrutiny over how easily users bypass safety measures. In their letter, the senators wrote: “We recognize that many companies maintain policies against non-consensual intimate imagery and sexual exploitation, and that many AI systems claim to block explicit pornography. In practice, however, as seen in the examples above, users are finding ways around these guardrails. Or these guardrails are failing.”

In response, X pointed to an update to Grok, the chatbot owned by Elon Musk’s xAI, which now prohibits edits of real people in revealing clothing and restricts image creation to paying subscribers. Meanwhile, a Reddit spokesperson stated that the platform prohibits non-consensual intimate media, does not offer tools to create it, and proactively removes it. Reddit previously took down a page displaying synthetic porn videos in 2018.

The regulatory landscape remains fragmented. In the U.S., the Take It Down Act—a federal law regarding nonconsensual sexualized imagery—became law in May, but its provisions focus primarily on individual users rather than holding image-generating platforms accountable. Consequently, state-level officials are stepping in; California’s attorney general recently opened an investigation into xAI’s chatbot. This probe follows statements from Musk claiming he was not aware of any naked underage images generated by Grok, despite xAI’s assertion that it removes illegal content, including CSAM and non-consensual nudity. The issue is further complicated by Chinese image and video generators, whose outputs spread to Western social platforms, and other AI models, such as Google’s AI, which generated racist videos that garnered millions of views.

Why it matters

The push from US lawmakers signals a shift from voluntary platform moderation to potential legislative mandates, as the proliferation of sexualized deepfakes outpaces current industry guardrails.