US deepfake crackdown takes effect as platforms face $53,000 fines
While major tech firms have pledged compliance, critics warn the legislation may be weaponised for political censorship and over-moderation
The US Take It Down Act, signed into law by President Donald Trump in May 2025, became fully enforceable on 19 May 2026. The legislation mandates that social media platforms remove nonconsensual intimate imagery, including AI-generated deepfakes, within 48 hours of notification. Failure to comply exposes platforms to civil penalties exceeding $53,000 per violation. The Federal Trade Commission has directed major technology firms, including Meta, Google, TikTok, and X, to adhere to these provisions.
The FTC sent letters to over a dozen tech companies ahead of the deadline, instructing them to offer users an easy takedown request process and remove offending content and any known identical copies. The agency reminded companies that violating the law could result in significant financial penalties. This enforcement action marks the transition from the law’s criminal provisions, which began with a conviction in Ohio, to its sweeping takedown requirements.
Major platforms have publicly supported the measure, citing alignment with existing safety protocols. Meta stated it had long fought intimate image abuse and was already compliant, while Snap noted its efforts align with ongoing safety investments. TikTok highlighted its zero-tolerance policy and partnerships with abuse prevention groups. Even X, which has faced scrutiny over sexually explicit AI imagery, has endorsed the law despite its history with such content.
However, the legislation has alarmed free speech advocates and abuse survivors. Mary Anne Franks, president of the Cyber Civil Rights Initiative, expressed skepticism that the law would be enforced fairly, citing President Trump’s comments about using the bill for himself. Franks warned that the law could be wielded against platforms the administration views as problematic, potentially targeting unpopular sites or specific speech groups.
Critics argue that the takedown provision may encourage companies to over-moderate non-offending content to reduce risk. Groups such as the Electronic Frontier Foundation have warned of potential overreach and censorship. There are also uncertainties regarding whether AI tools count as creators of nonconsensual imagery and how the law applies to private image production. The immediate impact on internet safety remains a subject of intense debate.
