Ftc

FTC Starts Enforcing TAKE IT DOWN Act

Federal agency begins policing fast takedowns for nonconsensual and AI-generated intimate images

Federal agency begins policing fast takedowns for nonconsensual and AI-generated intimate images

An illustration depicts a central artificial intelligence hub directing a flow of digital images into a folder labeled for takedown. © The GPU Trade Inc 2026


The Federal Trade Commission announced on May 19, 2026, that it has begun enforcing the TAKE IT DOWN Act, a law that requires covered online platforms to remove intimate photos and videos shared without consent when victims submit valid removal requests.

To be clear, Section 3 of the law set a one‑year compliance window and requires platforms to accept removal notices and take down the identified content and known identical copies within 48 hours of a valid request. The FTC is now using its enforcement authority to monitor and investigate compliance.

The agency also launched TakeItDown.ftc.gov, a public portal where victims or their authorized representatives can report platforms that fail to act or that lack any clear notice-and-removal process. The site is intended to feed the FTC’s investigations and help the agency prioritize enforcement.

Civil penalties are part of the enforcement toolbox. The FTC’s guidance and accompanying press materials remind businesses that violating the law may expose platforms to per‑violation fines — the agency has cited a penalty figure in the tens of thousands of dollars per missed takedown.

The statute explicitly covers nonconsensual intimate imagery whether it is authentic or a “digital forgery” created with software or artificial intelligence. That means AI‑generated deepfakes are squarely inside the law’s definition of the harms it targets.

In advance of enforcement the FTC sent letters to major tech companies reminding them of their obligations; the agency named large platforms and cloud services as targets of its outreach and said it expects systems to be in place no later than the May 19, 2026 deadline. The letters underscore that compliance is now an operational necessity, not just policy guidance.

The law’s arrival has particular salience for AI tools and services that can generate or amplify intimate images. Platforms that host model inference, run agent workflows, or provide image‑generation APIs may suddenly face the same notice‑and‑remove obligations as social networks if they distribute or publish NCII. That risk comes from the law’s broad focus on services that publish, host, or distribute such imagery.

Putting the takedown rules into practice is technically demanding. The FTC and outside analysts warn that a compliant removal must propagate across primary databases, CDNs, thumbnails, backups, and mirrored storage within 48 hours, and that AI‑native content often requires perceptual similarity matching rather than simple file hashing. Platforms face real engineering and product tradeoffs to meet the timetable.

Civil liberties groups and tech policy experts say the law raises free‑speech and misuse risks. Organizations including the Electronic Frontier Foundation and the Center for Democracy & Technology warned during debate and after passage that a fast takedown regime with limited appeal tools could be abused by bad‑faith reporters and chill lawful expression. Those concerns have not disappeared now that enforcement has begun.

The FTC’s enforcement model is administrative rather than private. Victims cannot sue platforms directly under the law; instead the statute vests the FTC with exclusive civil enforcement authority and asks the agency to bring actions when platforms fail to comply. The agency’s blog and guidance also lay out procedural expectations, including confirmation numbers and status tracking for removal requests.

What should platforms and AI developers do right now? The FTC’s business guidance recommends a clear, conspicuous reporting mechanism; rapid incident workflows; logging that proves timely action; and technical measures like perceptual hashing, similarity matching, and cooperation with resource centers that track NCII hashes. Legal and product teams should also document policies that guard against abuse of the complaint pipeline.

The TAKE IT DOWN Act’s enforcement phase will test how regulators and platforms balance rapid victim relief, engineering realities, and constitutional limits. Expect legal challenges, more detailed agency guidance, and iterative changes to platform workflows as cases and complaints surface the policy’s frictions and edge cases. For now, the FTC is signaling that fast takedowns are the new baseline for platforms in an AI age.