TL;DR
The TAKE IT DOWN Act is a federal law, signed May 19, 2025, that covers real non-consensual intimate imagery and AI-generated digital forgeries, which DMCA cannot reach. It makes knowingly publishing such imagery a federal crime and requires covered platforms to remove reported content, and known identical copies, within 48 hours of a valid notice. A valid notice needs a signature, enough information to locate the content, a good-faith statement, and contact details; the FTC enforces platform compliance.
The TAKE IT DOWN Act (Public Law 119-12) is a federal law, signed on May 19, 2025, that creates a legal pathway for removing non-consensual intimate imagery, including images and videos that were never real. It responds to two overlapping problems: intimate images shared online without the depicted person's consent, and AI tools that can generate convincing fake intimate imagery of a real, identifiable person who never posed for anything. Before this law, the second problem had no clean legal remedy. This piece covers what the Act covers, how the removal process works, and what it requires of the platforms that handle reports.
What counts as covered content
The statute defines two categories of covered content and treats them the same way procedurally, even though they arise very differently.
The first is authentic non-consensual intimate imagery: real photos or videos taken or shared without the depicted person's consent, sometimes called "revenge porn" in earlier state-law contexts. Most states already had some kind of criminal statute on this before the federal law passed.
The second is what the statute calls a "digital forgery": intimate content generated or manipulated using software (including AI) to depict a real, identifiable person in a way that did not actually happen. This is the category that had no workable legal remedy before 2025. A deepfake doesn't infringe anyone's copyright, because nothing was copied from an original work. It's a fabrication, not a reproduction. The Act covers both minors and adults, though the criminal provisions differ by the age of the depicted person, with more severe penalties when the depicted individual is a minor.
Why existing law didn't cover this
The takedown mechanism most commonly used before this law, a DMCA notice under 17 U.S.C. §512, is copyright law. It lets the creator of an original work demand removal of unauthorized copies of that work. That fits a genuine photo or video that was taken with the subject's knowledge and later leaked or reposted without permission: something real was copied without authorization, and DMCA is built for exactly that.
It fits a deepfake very poorly. Nothing was copied. The image or video was fabricated from scratch using a real person's likeness as an input, not lifted from an existing file. A platform that receives a DMCA notice about AI-generated content is being asked to apply a copyright remedy to something that isn't a copyright violation, and notices filed this way are routinely rejected or stall for that reason. The harm to the depicted person is often the same in both cases, but the legal basis for removal is different. That is the gap the TAKE IT DOWN Act was written to close.
What the Act actually does
The law creates two separate obligations, which fall on different actors.
A federal criminal offense. Knowingly publishing covered non-consensual intimate imagery of an identifiable adult, or of a minor, is now a federal crime. Penalties scale with the severity of the offense and are more severe when the victim is a minor. This provision targets the person who posts the content.
A platform takedown mandate. The law also requires "covered platforms" (general-purpose social media, content-hosting, and similar online services that meet certain size and functionality thresholds) to set up a process for receiving removal requests and to act on a valid request within a defined timeframe. This provision targets the platform hosting the content, not the original poster, and it's the one most people deal with directly: you file a notice and get content taken down.
The notice-and-removal process
A defined removal window
A valid TAKE IT DOWN Act notice obligates a covered platform to remove the reported content, and make reasonable efforts to remove known identical copies, as soon as possible but no later than 48 hours after receipt. That's a statutory deadline, not a courtesy response time that varies by platform, as ordinary content-moderation reports or even DMCA processing often do.
A request can come from the depicted individual, from a parent or guardian if the depicted person is a minor, or from another authorized representative acting on that person's behalf. The statute explicitly permits an authorized third party to file.
For a notice to be valid and start the 48-hour clock, it has to include four elements:
- A physical or electronic signature of the person submitting the request (the depicted individual, or their authorized representative).
- Enough information for the platform to locate the content, typically the specific URL or URLs plus any identifying details needed to find it.
- A brief statement made in good faith that the content is non-consensual.
- Contact information sufficient for the platform to reach the person who submitted the request.
An incomplete notice doesn't start the clock. If any of the four elements is missing, the platform isn't yet obligated to act within 48 hours. A common reason self-filed requests stall is that one element was left out or was too vague to act on, such as a URL without enough context to locate the content, or a signature block missing contact information.
Not sure your situation qualifies?
Run it through the TAKE IT DOWN Act eligibility checker — a short wizard that tells you whether covered content is likely to qualify before you spend time drafting a notice.
Enforcement: what happens if a platform ignores a valid notice
A covered platform that fails to comply with a valid, complete notice is treated as having engaged in an unfair or deceptive act or practice under the FTC Act, the legal category the Federal Trade Commission uses against a wide range of consumer-protection violations. That gives the FTC direct enforcement authority over noncompliant platforms, backed by civil penalties that are periodically adjusted for inflation and scale per violation.
The statute also gives platforms a good-faith safe harbor. A platform that removes content in response to a facially valid notice isn't liable for the removal, even if the content later turns out to fall outside the statute's scope. That is a large part of why platforms tend to act on notices that look complete rather than investigating each one first: the legal risk of over-removing in good faith is deliberately lower than the risk of under-removing.
Which platforms are covered
The duty to build a takedown process and honor the 48-hour window applies to "covered platforms" as the statute defines them: generally, online services and applications that primarily provide a forum for user-generated content, and that meet size thresholds tied to user base and revenue. The aim is to capture mainstream social platforms, image- and video-hosting services, and similar consumer-facing services where this kind of content typically circulates, not every website on the internet. A small or narrowly scoped site below those thresholds may fall outside the platform-obligation provisions, although the criminal prohibition on knowingly publishing covered content still applies to whoever posted it.
How it relates to copyright takedowns
The TAKE IT DOWN Act and DMCA address different situations and aren't interchangeable, but both can be relevant to the same broader problem at different moments. Genuine content that was authorized when it was created and later shared without permission is a copyright matter, and DMCA remains the right tool. Fabricated content, such as a deepfake, a face-swap, or an AI-generated composite depicting someone who never appeared in the source material, isn't a copyright violation at all and has to be filed under the TAKE IT DOWN Act. Filing under the wrong statute is one of the more common reasons a self-filed removal request is rejected or ignored, so working out which category applies matters as much as filing the notice.
In practical terms, the Act filled a real gap. Before it, fabricated intimate imagery of a real person had a criminal dimension in many states but no reliable, fast, federally mandated platform-removal mechanism. Now there is a defined process, a hard 48-hour deadline, and FTC enforcement behind it.
If you're weighing this against DMCA, a civil suit, or a plain platform report for a specific piece of content, the legal options decision tree walks through which applies. And because state law didn't disappear when this federal law arrived, the deepfake law map shows what additional state-level protections might apply on top of it.