TAKE IT DOWN Act
Last reviewed
Sources
18 citations
Review status
Source-backed
Revision
v1 ยท 2,028 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Sources
18 citations
Review status
Source-backed
Revision
v1 ยท 2,028 words
Add missing citations, update stale details, or suggest a clearer explanation.
The TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act) is a United States federal law that criminalizes the nonconsensual publication of intimate imagery, including AI-generated deepfake content, and requires online platforms to remove such material within 48 hours of a valid request from a victim. Sponsored by Senators Ted Cruz (Republican of Texas) and Amy Klobuchar (Democrat of Minnesota), the bill passed both chambers of Congress with broad bipartisan support and was signed into law by President Donald Trump on May 19, 2025, as Public Law 119-12. First Lady Melania Trump championed the legislation as a signature priority of her "Be Best" initiative.[1][2][3] The law has two distinct components: criminal prohibitions that took effect on enactment and are enforced by the Department of Justice, and a notice-and-takedown mandate for "covered platforms" enforced by the Federal Trade Commission (FTC), which became binding one year later on May 19, 2026.[4][5][6]
The Act is one of the first federal statutes in the United States to directly target AI-generated nonconsensual intimate imagery. It was widely praised by victims' advocates as a long-sought federal criminal remedy, while civil-liberties groups including the Electronic Frontier Foundation (EFF) and the Cyber Civil Rights Initiative (CCRI) warned that its takedown system is vulnerable to abuse and could suppress lawful speech.[7][8][9]
Nonconsensual intimate imagery (NCII), sometimes called "revenge porn," refers to the distribution of sexually explicit or intimate images of a person without that person's consent. Before the TAKE IT DOWN Act, the United States had no comprehensive federal criminal statute covering NCII; enforcement depended on a patchwork of state laws that varied widely in scope and penalties. A 2022 reauthorization of the Violence Against Women Act created a federal civil cause of action for victims, codified at 15 U.S.C. 6851, but it provided no criminal penalty.[7][10]
The rapid spread of consumer generative AI tools sharpened the problem. Inexpensive "nudify" applications and image models made it possible to fabricate convincing explicit depictions of real people, including minors, from ordinary photographs. High-profile incidents, including AI-generated explicit images of the singer Taylor Swift that circulated on the platform X in January 2024, drew national attention and added political momentum behind a federal response. The TAKE IT DOWN Act addresses both authentic images and "digital forgeries," which the statute defines as intimate visual depictions created through machine learning, artificial intelligence, or other technological means that a reasonable person would find indistinguishable from an authentic depiction of the individual.[1][7][10]
The law is structured in two operative parts. Section 2 amends Section 223 of the Communications Act of 1934 (47 U.S.C. 223) to create new federal crimes for knowingly publishing, or threatening to publish, NCII of identifiable individuals, whether the images are authentic or digital forgeries. Consent to the creation of an image does not constitute consent to its publication. The statute includes exceptions for law-enforcement activity, legal proceedings, legitimate medical or educational purposes, disclosures reasonably intended to help the depicted individual, and matters of legitimate public concern.[3][4][10]
Section 3 imposes a notice-and-removal duty on "covered platforms." A covered platform is a public-facing website, online service, or application that primarily provides a forum for user-generated content, or that in the regular course of business publishes or makes available NCII. Broadband internet access providers, email services, and services consisting primarily of preselected (non-user-generated) content are excluded. Covered platforms must establish a clear and accessible process for victims or their representatives to request removal, take down reported material within 48 hours of a valid request, and make reasonable efforts to identify and remove known identical copies. Platforms that remove content in good faith based on a facially valid request receive liability protection, regardless of whether the material ultimately proves unlawful.[4][5][6][11]
The criminal penalty structure distinguishes among categories of conduct.
| Offense category | Maximum prison term |
|---|---|
| Publishing authentic NCII of an adult | 2 years |
| Publishing authentic NCII of a minor | 3 years |
| Publishing a digital forgery (deepfake) of an adult | 2 years |
| Publishing a digital forgery (deepfake) of a minor | 3 years |
| Threatening to publish NCII of an adult | 18 months |
| Threatening to publish NCII of a minor | 30 months |
Each offense also carries fines. The Act includes a rule of construction stating that it does not limit the application of other laws, including the federal child sexual abuse material statute at 18 U.S.C. 2252.[4][10][12]
Senator Cruz first introduced the TAKE IT DOWN Act as S.4569 in the 118th Congress in June 2024. The Senate passed that version by unanimous consent on December 3, 2024, but the House did not act before the session ended. Cruz and Klobuchar reintroduced the measure as S.146 in January 2025, with a companion bill, H.R.633, introduced in the House. The Senate again passed the bill by unanimous consent, on February 13, 2025. The House passed it on April 28, 2025, by a vote of 409 to 2, with Representatives Thomas Massie (Republican of Kentucky) and Eric Burlison (Republican of Missouri) casting the only opposing votes. President Trump signed it into law on May 19, 2025, in a White House ceremony at which Melania Trump symbolically added her own signature.[1][2][13][14]
| Date | Event |
|---|---|
| June 2024 | Cruz introduces S.4569 (118th Congress) |
| December 3, 2024 | Senate passes S.4569 by unanimous consent; House does not act |
| January 2025 | Cruz and Klobuchar reintroduce as S.146; H.R.633 introduced in House |
| February 13, 2025 | Senate passes S.146 by unanimous consent |
| April 28, 2025 | House passes 409-2 |
| May 19, 2025 | Signed into law as Public Law 119-12; criminal provisions effective |
| May 19, 2026 | Platform notice-and-takedown duty becomes enforceable |
Enforcement is split between two effective dates, a distinction central to how the law operates. The criminal prohibitions in Section 2 took effect immediately upon enactment in May 2025 and are prosecuted by the Department of Justice. The platform obligations in Section 3 were given a one-year runway: covered platforms had until May 19, 2026, to stand up a removal process and begin honoring the 48-hour rule. A platform's failure to reasonably comply is treated as a violation of a rule defining an unfair or deceptive act or practice under the FTC Act, allowing the Commission to seek civil penalties of up to $53,088 per violation.[4][5][6]
The FTC began enforcing Section 3 on May 19, 2026. In the days before the deadline, FTC Chairman Andrew Ferguson sent warning letters to a group of major platforms, including Alphabet, Amazon, Apple, Automattic, Bumble, Discord, Match Group, Meta, Microsoft, Pinterest, Reddit, SmugMug, Snapchat, TikTok, and X, urging them to come into full compliance.[5][6]
The first criminal cases arrived in 2026. On April 7, 2026, James Strahler II, of Columbus, Ohio, became the first person convicted under the Act after pleading guilty in federal court to charges that included publication of digital forgeries, alongside cyberstalking and production of child sexual abuse material. Prosecutors said Strahler used more than two dozen AI applications to fabricate explicit images of at least six adult victims and of minors between December 2024 and June 2025, and that he harassed victims by sending the material to their associates. Melania Trump publicly hailed the conviction. In late May 2026, around the law's first anniversary, federal prosecutors in the Eastern District of New York charged two additional men, Cornelius Shannon and Arturo Hernandez, with producing thousands of deepfake images depicting celebrities, public officials, and acquaintances.[15][16][17]
The Act drew sustained opposition from digital-rights and free-expression advocates even as it advanced toward near-unanimous passage. The EFF argued that the notice-and-takedown regime threatens free expression, privacy, and due process without effectively helping victims. Its central objections were that the removal provision sweeps more broadly than the law's own NCII definitions, potentially reaching any intimate or sexual content; that the 48-hour window is too short for platforms to verify whether a given report is legitimate; and that providers, especially smaller ones, would respond by removing reported material first and asking questions later to avoid legal risk. The EFF also warned that the law lacks clear exemptions for end-to-end encrypted services, which by design cannot inspect user content, creating pressure to weaken or abandon encryption.[7][8]
The Cyber Civil Rights Initiative, a leading anti-NCII advocacy group, supported a federal criminal prohibition but opposed the takedown provision as enacted, calling it unconstitutionally vague and overbroad and lacking adequate safeguards against misuse. CCRI cautioned that bad-faith actors, disgruntled individuals, or commercial competitors could weaponize the reporting process to remove lawful content they merely disapprove of, and criticized an exception for a person who possesses or publishes an intimate depiction of himself or herself as a loophole that could permit nonconsensual disclosure when the discloser also appears in the image.[9]
Critics across the political spectrum also raised concerns about selective enforcement. Some commentators noted that President Trump had spoken about using the law against content directed at himself, which civil-liberties groups cited as illustrating the risk that a broad takedown power could be turned against protected speech such as satire, journalism, or political commentary.[8][18]
The TAKE IT DOWN Act marks a significant shift in United States internet and AI regulation. It is among the first federal laws to impose affirmative content-removal duties on a wide range of online platforms and to do so through an FTC-enforced unfair-practices framework rather than through changes to platform liability shields such as Section 230 of the Communications Decency Act. By explicitly covering "digital forgeries," it is also one of the earliest federal statutes written to address synthetic media produced by generative AI, and it has been cited as a template in debates over related proposals such as the NO FAKES Act and the DEFIANCE Act.[1][11][18]
The law's long-term impact remains contested. Supporters point to its rapid, near-unanimous passage and the first 2026 convictions as evidence of a workable federal tool against image-based sexual abuse. Skeptics argue that the broad takedown mandate, the absence of penalties for false reports, and the unresolved questions around encrypted services could generate over-removal of lawful speech and constitutional challenges as enforcement matures.[7][9][11]