The legal question people actually search is some version of: is deepfake porn illegal? The honest 2026 answer: making or sharing it of a real person without consent is unlawful in most of the developed world — and where the subject is a minor it is prosecuted as child sexual abuse material, full stop. This page maps the law by jurisdiction and lays out what victims and bystanders can do.
Not legal advice. Statutes change and facts matter. If you are facing a real situation — as victim, accused, or platform — get a lawyer in your jurisdiction. What follows is an orientation map, not counsel.
1.The three legal questions, separated
Confusion about deepfake porn legality usually comes from conflating three distinct questions that different statutes answer differently:
Is creating it a crime? Increasingly yes — the UK criminalized creation itself in 2024; several US states criminalize creation with intent to distribute or harass; elsewhere creation alone may be a civil rather than criminal wrong.
Is sharing it a crime? Almost everywhere now. Distribution without consent is the act nearly every statute targets first, because distribution is where the harm becomes irreversible.
Is possessing or viewing it a crime? For adult subjects, usually not — possession statutes remain rare. For subjects who are or appear to be minors, possession is a serious felony in essentially every jurisdiction, with no "it's AI" defense.
2.United States: federal floor + state patchwork
The federal layer
The Take It Down Act (enacted May 2025) is the first US federal statute squarely covering synthetic intimate imagery. Its two operative parts:
Criminal provision: knowingly publishing (or threatening to publish) non-consensual intimate imagery — expressly including AI-generated depictions — carries fines and up to two years' imprisonment, three where the depicted person is a minor.
Platform provision: covered platforms must operate a notice-and-removal process and take down flagged NCII within 48 hours of a victim's verified request, plus make reasonable efforts to remove identical copies.
Synthetic CSAM doesn't need this statute — it routes into the existing child-exploitation framework (18 U.S.C. §1466A and state analogues), which courts have applied to AI-generated material depicting identifiable or fictional minors.
The state layer
At least 45 states have enacted statutes touching synthetic intimate imagery since the 2019–2024 wave. The patterns:
Pattern
What it does
Examples
Criminal distribution statutes
Misdemeanor-to-felony penalties for sharing AI-NCII without consent
California, Texas, Florida, Virginia, Georgia and most others
Creation offenses
Criminalize making the imagery itself, usually with intent element
Subsets of the above; expanding each session
Civil causes of action
Private right to sue for damages, often with statutory minimums and attorney fees
California AB-602-lineage statutes; NY civil remedies
Election-specific laws
Separate deepfake rules for political content (not intimate imagery)
~20 states — adjacent but distinct regime
The practical point for US readers: the victim's home state law usually applies to harm suffered there, and civil suits can proceed regardless of whether a prosecutor acts. The Cyber Civil Rights Initiative maintains the canonical 50-state tracker.
3.United Kingdom: the strictest mainstream regime
The UK moved furthest and fastest. Under the Online Safety Act 2023, sharing intimate images — including deepfakes — without consent is a criminal offense. The 2024 amendment closed the creation loophole: making a sexually explicit deepfake of an adult without consent is itself an offense, even if never shared. OFCOM's platform regime adds proactive duties — services must prevent NCII, not merely respond to reports — with fines up to 10% of global turnover.
The UK position thus answers all three questions above with "yes, criminal": creating, sharing, and (for minors under pre-existing law) possessing.
4.European Union: three converging instruments
Directive (EU) 2024/1385 on violence against women requires member states to criminalize non-consensual production and sharing of intimate images, explicitly covering manipulated material. National transposition is rolling through 2025–2027.
The AI Act (2024/1689) doesn't criminalize content but mandates marking AI-generated material — making undisclosed synthetic imagery a regulatory violation on top of content-law exposure.
GDPR provides the privacy-law layer: a recognizable likeness is personal data; processing someone's photos to synthesize nudes has no lawful basis, exposing operators and uploaders to data-protection enforcement alongside criminal law.
Member-state law often predates the EU instruments — Germany, France, Denmark, and others criminalized synthetic NCII earlier under image-rights and sexual-offense statutes.
5.Rest of world, briefly
Jurisdiction
Status (Nov 2026)
South Korea
Production AND possession criminalized; up to 7 years; post-2024 enforcement is aggressive and publicized
Australia
Criminal offense to create/share sexual deepfakes without consent; eSafety takedown powers
Canada
Covered by intimate-image and deepfake provisions; provincial civil remedies
India
IT Act/IT Rules duties + IPC provisions; rapid-removal obligations on intermediaries
Japan
Prefectural statutes + 2024 national legislation tightening deepfake sexual imagery
Offshore havens
Where the infrastructure hides — weak enforcement, not legal permission; operators there still face extradition exposure and infrastructure bans
6.The risk ledger for uploaders and buyers
People who use undress tools typically price in "the site might scam me" and ignore the larger ledger. For completeness:
Criminal exposure as above — in the strictest jurisdictions the act of uploading the photo is already the offense.
Data exposure — the uploaded photo of your victim goes to an anonymous operator with no incentive to protect it; several tool databases have already leaked, exposing both the victims' photos and the users' payment/email trails.
Payment trace — card transactions and even "anonymous" crypto payments are routinely traceable by investigators; Telegram Stars purchases leave a platform-side record.
7.What actually works for victims
Preserve evidence before reporting. URLs, timestamps, screenshots of the content in context. Takedown destroys the record; preserve first.
Use the platform's NCII report channel. Covered platforms must act within 48 hours under Take It Down; major platforms had equivalent policies earlier.
Fingerprint via StopNCII (adults) or Take It Down (minors). Hash-matching blocks re-uploads across participating platforms without the image ever leaving your device. Highest-leverage single step.
De-index from search engines. Google and Bing both run dedicated synthetic-NCII removal forms. Search removal cuts real-world exposure faster than chasing sites through domain churn.
File a police report / consult counsel. Post-2024, prosecutors bring these cases; civil statutes provide damages where criminal proof is harder.
Use specialist support. Cyber Civil Rights Initiative (US), Revenge Porn Helpline (UK), national equivalents — free, experienced, and they treat synthetic abuse identically to photographic abuse.
The single most important thing a victim should know: you are not required to prove the image is fake to get help. The harm is the non-consensual intimate depiction of you — its synthetic origin is a technical detail, not a defense.