Living evidence map · scoping-review idiom
TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act)
US-TAKEITDOWN-2025 · US
The TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act), Public Law 119-12 (139 Stat. 55), signed May 19, 2025, is one of the few binding federal AI-specific statutes in the United States. It has two operative halves. First, it criminalizes the knowing publication of nonconsensual intimate visual depictions of identifiable adults (obtained under a reasonable expectation of privacy and intended to cause, or causing, harm) and of minors (under a stricter intent standard), and it expressly reaches AI-generated 'digital forgeries' — intimate depictions created through software, machine learning, or artificial intelligence that are indistinguishable from authentic images; four of its seven offenses are deepfake-specific, with penalties up to two years' imprisonment (adults) or three years (minors) plus mandatory restitution and forfeiture. Second, it requires 'covered platforms' (user-generated-content websites, online services, and applications) to establish a notice-and-removal process and remove a reported nonconsensual intimate depiction — including a deepfake — within 48 hours of a valid request; platforms had until May 19, 2026 to implement the process. Non-compliance is enforced by the Federal Trade Commission as an unfair or deceptive act or practice under the FTC Act; there is no private right of action. The Act is deliberately takedown-focused — it imposes no watermarking, labeling, or content-provenance duty.
Background & scope
TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act) addresses 1 contested AI-governance topic explicitly, 1 via general principles.
Provisions & coverage
- governsDeepfakes / Synthetic ContentPub. L. 119-12 — criminalizes nonconsensual intimate 'digital forgeries' (AI deepfakes) of adults and minors and requires covered platforms to remove them within 48 hours; the statute names 'artificial intelligence' in its operative digital-forgery definition[1]
- implicitIndividual RedressPub. L. 119-12 — the 48-hour platform notice-and-removal process plus mandatory criminal restitution and forfeiture give nonconsensual-intimate-image / deepfake victims a targeted remedy; narrow to one harm domain and FTC-enforced with no private right of action[1]
Enforcement & impact
Cross-jurisdiction comparison
How peer instruments treat the topics TAKE IT DOWN Act (Tools to Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act) governs.
| Topic | EU-AIA-2024 | US-EO-14110 | US-EO-14179 | UK-WHITEPAPER-2023 | CN-GENAI-2023 | G7-HIROSHIMA | OECD-AI-PRIN | COE-AI-CONV | UN-RES-2024 | NIST-AI-RMF | BLETCHLEY-2023 | SEOUL-2024 | NIST-AI-RMF-GENAI | CA-SB-1047 | IN-DPDP-2023 | BR-AIBILL-2024 | ASEAN-AI-GUIDE-2024 | AU-AI-STRATEGY-2024 | ANTHROPIC-RSP-2024° | OPENAI-PREPAREDNESS-2023° | DEEPMIND-FSF-2024° | META-FRONTIER-2024° | UK-US-AISI-MOU-2024 | WH-VOLUNTARY-2023 | SG-MODEL-AI-2024 | JP-METI-AI-2024 | EU-GDPR-2016 | EU-GPAI-COP-2025 | OMB-M-24-10 | GSA-AI-GUIDE-2024 | DOD-RAI-2022 | FEDRAMP-AI-2024 | DFARS-252-204 | CA-SB-53 | CA-SB-243 | CA-SB-942 | EU-PLD-2024 | UNESCO-AI-ETHICS-2021 | EU-PWD-2024 | CN-DEEPSYN-2022 | NY-RAISE-2025 | IT-AILAW-2025 | JP-AIPROMO-2025 | UN-GDC-2024 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Deepfakes / Synthetic Content | governs | governs | silent | silent | governs | governs | silent | silent | implicit | implicit | silent | silent | governs | silent | governs | silent | silent | silent | silent | silent | silent | silent | silent | governs | governs | silent | silent | silent | silent | silent | silent | silent | silent | silent | silent | implicit | silent | silent | silent | governs | silent | governs | silent | silent |
°= industry self-imposed voluntary framework. Comparing a voluntary code's "governs" tint with a binding regulation's "governs" tint flattens the legal-force distinction; use the instrument-page banner for the operative status of each.
See also
Per-audience views
- Provisions →Article-by-article obligation breakdown for procurement + RFP authors.
- Disclosure form →Vendor-disclosure questionnaire derived from this instrument's operative obligations.
- Harm narratives →Documented harms relevant to this instrument's topics, for civil-society advocacy.
- Briefing pack →Journalist-ready summary with quotes + dates + primary-source links.
Article tools — track changes, suggest an edit
View history — every captured revision of this article · What links here
Further reading
32 academic & grey-literature sources on the topics this instrument addresses (not commentary on the instrument itself) — catalogued metadata with a primary link; one-line findings are ✦ AI-generated summaries, labeled as such (charter §7.9). Browse the full literature index.
- The Current Landscape of Deepfake Legislation in the United States Peer-reviewed✦ AIThematic analysis of 319 state deepfake bills (2019-2024) finds a fragmented patchwork concentrated on political and sexually-explicit content.
- Reimagining U.S. Tort Law for Deepfake Harms: Comparative Insights from China and Singapore Peer-reviewed✦ AIArgues fragmented US tort doctrines (defamation, publicity, IIED) are ill-suited to deepfake harms and draws remedial lessons from Chinese and Singaporean law.
- A Teleological Interpretation of the Definition of DeepFakes in the EU Artificial Intelligence Act—A Purpose-Based Approach to Potential Problems With the Word 'Existing' Peer-reviewed✦ AIWarns a narrow reading of 'existing' in the AI Act's deepfake definition could exclude synthetic media from transparency duties, urging a teleological interpretation.
- Audio deepfakes and the regulation of the landlords of creativity Peer-reviewed✦ AIArgues US, EU and Chinese regimes fail to assign audio-deepfake liability to 'landlords of creativity' (foundation-model providers) and proposes holding them accountable.
- Identifying Algorithmic Decision Subjects' Needs for Meaningful Contestability Peer-reviewed✦ AIEmpirically elicits what decision subjects need for contestation to be 'meaningful', informing the design of effective remedies and appeal mechanisms for ADM.
- Two Means to an End Goal: Connecting Explainability and Contestability in the Regulation of Public Sector AI Preprint✦ AIInterview study with 14 regulation experts distinguishes judicial vs non-judicial and individual vs collective contestation channels for public-sector AI remedies.
- Human detection of political speech deepfakes across transcripts, audio, and video Peer-reviewed✦ AIExperiments show "audio and visual information enables more accurate discernment than text alone" — humans rely more on how something is said than on transcript content.
- When non-consensual intimate deepfakes go viral: The insufficiency of the UK Online Safety Act Peer-reviewed✦ AIArgues the UK Online Safety Act 2023 inadequately addresses non-consensual intimate deepfakes as image-based sexual abuse, leaving enforcement and takedown gaps.
- AI or Your Lying Eyes: Some Shortcomings of Artificially Intelligent Deepfake Detectors Peer-reviewed✦ AIArgues detector-based solutions depend on scarce institutional trust and risk undermining epistemic autonomy, so purely technological fixes for deepfakes are dim.
- The Liar's Dividend: Can Politicians Claim Misinformation to Evade Accountability? Peer-reviewed✦ AIFive survey experiments (>15,000 US adults) show false 'it's a deepfake/fake news' claims can help politicians retain support, evidencing the liar's dividend.
- Deep fakes and the Artificial Intelligence Act—An important signal or a missed opportunity? Peer-reviewed✦ AICritiques the EU AI Act's placement of deepfakes in the 'limited risk' tier, leaving transparency obligations as the only direct safeguard without bans or victim remedies.
- Non-Consensual Synthetic Intimate Imagery: Prevalence, Attitudes, and Knowledge in 10 Countries Peer-reviewed✦ AISurvey of >16,000 respondents across 10 countries finds NSII victimization/perpetration persists even where specific laws exist, suggesting current laws under-deter.
+ 20 more across this instrument's topics — see the literature index.
References
The primary instrument sources behind the article's classifications.
- TAKE IT DOWN Act, Pub. L. No. 119-12, 139 Stat. 55 (2025) (platform notice-and-removal at 47 U.S.C. § 223 / § 223a note (Communications Act of 1934 § 223), FTC-enforced under the FTC Act (15 U.S.C. § 57a); criminal provisions at 18 U.S.C. §§ 2252, 2256, 2264; the borrowed 'intimate visual depiction' definition is from 15 U.S.C. § 6851)
- Pub. L. 119-12 — criminalizes nonconsensual intimate 'digital forgeries' (AI deepfakes) of adults and minors and requires covered platforms to remove them within 48 hours; the statute names 'artificial intelligence' in its operative digital-forgery definition
- Pub. L. 119-12 — the 48-hour platform notice-and-removal process plus mandatory criminal restitution and forfeiture give nonconsensual-intimate-image / deepfake victims a targeted remedy; narrow to one harm domain and FTC-enforced with no private right of action
How to cite this article
Cite this article
8 formats · 1-click copyPersistent identifier: https://policywindow.org/wiki/take-it-down-act — committed-stable URL with content-versioning via ?asOf= (rollout pending per methodology §7). DOIs via Zenodo are on the roadmap.
Does this instrument’s approach work? — the social-science evidence
Aggregated over the 2 topics this instrument governs: whether each harm is empirically real, and whether the peer-reviewed evidence shows governance reduces it. The badge is the epistemic status of the evidence— “thin”/“absent” efficacy evidence is itself a finding (the “second silence”). Each epistemic-status label is Policy Window's editorial assessment of the cited evidence base (a structured classification), not a verdict any single source issues.
Of the 2 governed topics with a social-science evidence review, evidence that governance reduces the harm is established for 0, contested for 0, thin for 1, and absent for 1 — for most, no replicated study yet shows this instrument's approach works (the "second silence").
Deepfakes / Synthetic Content
The flagship harm — non-consensual sexual deepfakes — is empirically real and sharply gendered: content audits find ~96-98% of deepfake videos online are non-consensual pornography overwhelmingly depicting women, and a pre-registered 10-country survey (>16,000 people) found 2.2% reporting victimization and 1.8% perpetration of synthetic intimate imagery, with documented mental-health, career, and participation harms. By contrast, the parallel claim that political/informational deepfakes UNIQUELY deceive is contested-to-refuted: experiments find deepfakes about as (not more) credible than equivalent text/audio fakes, and a 56-paper meta-analysis (k=137, N=86,155) puts unaided human detection near chance — implying a detection problem more than an exceptional-persuasion one.
Sources: Umbach, Henry, Beard & Berryessa 2024 (CHI '24, 'Non-Consensual Synthetic Intimate Imagery ... in 10 Countries'); Diel et al. 2024 (Computers in Human Behavior Reports 16:100538, deepfake-detection meta-analysis of 56 papers); Barari, Lucas & Munger 2025 (Journal of Politics 87(2), 'Political Deepfakes Are as Credible as Other Fake Media'); Flynn et al. 2022 (British Journal of Criminology, multi-country image-based sexual abuse study)
Direct impact evidence that deepfake governance reduces the targeted harm is sparse and, where it exists, discouraging: the one quasi-experimental evaluation (Cuevas & Horta Ribeiro 2025, synthetic-control across three platforms) found the U.S. TAKE IT DOWN Act's passage plus the MrDeepfakes shutdown did NOT suppress synthetic non-consensual imagery — posting rose above counterfactual baselines and displaced elsewhere. Technical enforcement is likewise unreliable: detectors fail to generalize to unseen generators (notably diffusion models) and are vulnerable to adversarial evasion, with in-the-wild accuracy well below benchmark figures. No rigorous evaluation yet shows a deepfake-specific law, takedown mandate, or watermarking scheme producing a sustained reduction in prevalence or harm.
Sources: Cuevas & Horta Ribeiro 2025 ('Deepfake Pornography is Resilient to Regulatory and Platform Shocks', arXiv:2602.02754); 'Adversarial Reality for Evading Deepfake Image Detectors' (ICCVW 2025); TAKE IT DOWN Act, S.146 / Pub. L. 119-12 (2025); CRS Legal Sidebar LSB11314
Individual Redress
The premise behind redress — that affected people lack meaningful recourse against automated decisions — is real, but the flagship instrument is weaker than commonly assumed. Wachter, Mittelstadt & Floridi (2017) show GDPR creates only a limited 'right to be informed,' not a binding 'right to explanation' of specific decisions; and controlled work finds the explanations actually delivered do not measurably improve lay decision accuracy over showing the bare AI prediction (Alufaisan et al. 2021; and a 2022 meta-analysis by Schemmer et al. — screening 393 articles down to 9 in the final analysis — reports 'no effect of explanations on users' performance compared to sole AI predictions,' even though XAI overall had a positive effect). Honest caveat: the legitimacy/dignity value of being heard is empirically well established in the procedural-justice tradition even where outcome accuracy is unchanged, so 'redress fails' depends on which aim is measured.
Sources: Wachter, Mittelstadt & Floridi 2017 (International Data Privacy Law 7(2):76); Alufaisan, Marusich, Bakdash, Zhou & Kantarcioglu 2021 (Proceedings of the AAAI Conference on AI 35(8):6618); Schemmer, Hemmer, Nitsche, Kühl & Vössing 2022 (AAAI/ACM AIES '22, meta-analysis)
There is no rigorous impact evaluation showing that mandated redress mechanisms (right-to-explanation, appeal, human-in-the-loop review) actually reduce erroneous or unfair automated decisions — the evidence that the rule works is itself missing. The closest experimental analogues are discouraging: explanations increase humans' acceptance of AI recommendations regardless of correctness (Bansal et al. 2021), and algorithm-in-the-loop oversight can introduce racial disparities and exhibit automation bias rather than reliably catching model errors (Green & Chen 2019). The procedural-justice literature (Tyler 1990; Lind & Tyler 1988) robustly supports a legitimacy and compliance benefit of fair process, but it measures perceived fairness, not reduction of the substantive decision harm redress is meant to cure.
Sources: Bansal, Wu, Zhou, Fok, Nushi, Kamar, Ribeiro & Weld 2021 (CHI '21); Green & Chen 2019 (Disparate Interactions, ACM FAT* '19); Tyler 1990 (Why People Obey the Law, Yale Univ. Press); Lind & Tyler 1988 (The Social Psychology of Procedural Justice, Plenum Press)