Are Deepfakes Illegal? What US Law Actually Says in 2026

Are deepfakes illegal? If you are asking because someone made a fake of you, or because you are worried something you created could get you in trouble, the honest answer is: it depends on the use, not the technology. US law does not ban deepfakes outright. It targets specific harmful uses, and those laws now have real teeth.

Disclaimer: This article is general legal information, not legal advice. Deepfake laws change quickly, and how they apply depends on your facts and your state. If you are facing a real situation, consult a licensed attorney.

Are deepfakes illegal? If you are asking because someone made a fake of you, or because you are worried something you created could get you in trouble, the honest answer is: it depends on the use, not the technology. US law does not ban deepfakes outright. It targets specific harmful uses, and those laws now have real teeth.

Quick answer: Deepfakes are not illegal by default in the United States. The law targets specific uses: non-consensual intimate images are now a federal crime under the TAKE IT DOWN Act, many states ban deceptive election deepfakes, and fraud or defamation with deepfakes is illegal under existing law. Parody and satire remain protected.

Are Deepfakes Illegal? The Short Answer

There is no federal law that says "deepfakes are illegal." Making a face-swap video of yourself, a labeled parody of a politician, or a consensual film effect breaks no law by itself.

What the law regulates is what the deepfake depicts and what you do with it. Three questions decide most cases:

1. What does it show? Intimate imagery without consent is treated most severely.
2. What is it used for? Deception, fraud, harassment, and election manipulation trigger liability.
3. Where did it happen? State laws vary widely, and 47 states now have some form of deepfake law on the books, according to Ballotpedia's 2026 legislative tracking.

The rest of this guide walks through each layer: federal law, state law, and the specific uses that cross the line.

Federal Deepfake Laws in 2026

The TAKE IT DOWN Act: Non-Consensual Intimate Images

The biggest change in deepfake law arrived in 2025. The TAKE IT DOWN Act, signed into law on May 19, 2025, makes it a federal crime to knowingly publish non-consensual intimate images, explicitly including AI-generated "digital forgeries" of real, identifiable people, per the Congressional Research Service summary on congress.gov.

The law has two parts:

• Criminal provisions, effective at signing, with penalties including fines and imprisonment for publishing NCII of adults and minors.
• Platform takedown duties. Covered platforms had until May 19, 2026 to offer a removal request process and to take down reported intimate images, and known copies, within 48 hours. The FTC began enforcing these obligations in May 2026.

In plain English: sexual deepfakes of real people without consent are now a federal crime, and platforms are legally required to remove them fast when victims report them.

FCC Rules: AI Voices in Robocalls

Voice deepfakes have their own federal rule. In February 2024, the FCC issued a declaratory ruling that AI-generated voices in robocalls are "artificial" under the Telephone Consumer Protection Act, effective immediately. That makes voice-cloned robocalls illegal without the called party's prior consent and gives state attorneys general a direct enforcement tool against scam calls using cloned voices.

Fraud, Identity Theft, and Existing Federal Statutes

Even where no deepfake-specific statute applies, existing federal law often does. Using a deepfake to commit wire fraud, identity theft, extortion, or impersonation of a government official is illegal because the underlying conduct is illegal. Prosecutors do not need the word "deepfake" in a statute to charge a deepfake-enabled scam. The documented losses are substantial; our deepfake statistics page collects the major cases and numbers.

State Deepfake Laws: A Patchwork

States moved faster than Congress, and the result is a patchwork. As of mid-2026, 47 states have enacted some form of deepfake law, with Alaska, Missouri, and Ohio the holdouts, per Ballotpedia. Coverage clusters in four categories:

Tennessee's ELVIS Act, effective July 1, 2024, deserves a special mention: it was the first state law protecting a person's voice from unauthorized AI replication, not just their image.

Penalties, definitions, and deadlines differ meaningfully by state. For the full breakdown, see our deepfake laws by state table.

Is It Illegal to Make a Deepfake of Someone?

This is the worried creator's question, so here is the careful answer: making a deepfake of someone is not automatically illegal in most states, but it can become illegal quickly depending on content, consent, and use.

Creating vs Sharing: Where Liability Usually Attaches

Most laws, including the federal TAKE IT DOWN Act, target publication and distribution rather than private creation. Sharing is where the harm happens, so sharing is where most liability attaches.

But do not treat that as a safe harbor. A growing number of state laws reach creation itself for sexual deepfakes, especially involving minors, and possession or creation of AI-generated sexual content depicting minors is prosecuted severely under federal and state child sexual abuse material laws. Creation with intent to distribute or defraud can also be charged as an attempt or conspiracy.

Consent Changes Everything

A deepfake made with the subject's informed consent, think film de-aging, accessibility voice banking, or a consenting friend's birthday gag, is generally lawful. Strip the consent away and the same technology can be a crime, a tort, or both. If you are creating anything depicting a real person, get consent in writing, and label the result as synthetic. Labeling does not cure an otherwise illegal use, but it defeats the deception element many statutes require.

When Deepfakes Are Clearly Illegal

Non-Consensual Intimate Imagery

Publishing sexual or intimate deepfakes of a real person without consent is a federal crime under the TAKE IT DOWN Act and is separately illegal in 46 states. This is the clearest line in all of deepfake law. If this has happened to you, you have both criminal and civil options, and platforms must remove the content within 48 hours of a valid request. Our deepfake harassment victim resources page walks through the steps.

Election Deception Within Statutory Windows

About 30 states ban materially deceptive political deepfakes, typically within a window before an election (often 30 to 90 days) and typically with exceptions for labeled content, parody, and news coverage. Texas SB 751 (2019) and California AB 730 (2019, since succeeded by AB 2655 and AB 2839) were early examples; the map has grown every session since. Details and pending litigation are covered in our election deepfake laws guide.

Fraud, Extortion, and Impersonation Scams

Using a deepfake to steal is illegal everywhere, full stop. Cloned-voice phone scams, fake executive video calls authorizing wire transfers, and deepfake celebrity investment ads all fall under existing fraud, theft, and extortion statutes, plus the FCC's robocall rule when phones are involved. The deepfake is the method; the fraud is the crime.

Defamation and False Light

A deepfake that shows a real person doing or saying something they never did can support a defamation or false light lawsuit if it harms their reputation and a reasonable viewer would take it as real. These are civil claims: the target sues for damages. Public figures must additionally prove the creator knew it was false or acted recklessly, which a fabricated video often demonstrates by its nature.

When Deepfakes Are Legal

Parody, Satire, and Commentary

The First Amendment protects parody and satire, including AI-generated parody of politicians and celebrities. Obvious jokes, labeled spoofs, and political commentary are generally lawful, and most state deepfake statutes write explicit exceptions for them. The protection thins as content becomes deceptive: a "parody" indistinguishable from reality, published to mislead, can lose the shield.

Consensual and Disclosed Uses

Film and TV effects, dubbing and translation, accessibility tools that restore a lost voice, training simulations, and any use where the depicted person consented are legal. Best practice across all of them is disclosure: label synthetic media as synthetic. It keeps protected uses clearly on the right side of statutes that hinge on deception.

Deepfake Laws Outside the US

Other jurisdictions regulate more broadly. The EU AI Act imposes transparency duties: deepfakes must generally be disclosed as artificially generated or manipulated, with obligations phasing in through 2025 and 2026. Our EU AI Act deepfake rules guide covers the details. The UK criminalizes sharing intimate deepfakes under the Online Safety Act framework, with further offenses targeting creation in subsequent legislation. China requires labeling of AI-generated synthetic content under its deep synthesis and AI labeling rules. If your audience is global, assume disclosure duties apply.

What to Do If You Are a Victim of an Illegal Deepfake

1. Preserve evidence first. Screenshot the content, the URL, the account, and the date before reporting it. Takedowns can destroy the proof you need later.
2. Report to the platform. For intimate images, covered platforms must remove reported NCII within 48 hours under the TAKE IT DOWN Act; the FTC publishes compliance guidance.
3. Use the dedicated takedown services. NCMEC's Take It Down service helps minors and adults remove intimate images; StopNCII.org helps adults hash and block images across participating platforms.
4. Report crimes. File with local police and the FBI's Internet Crime Complaint Center at ic3.gov for fraud, extortion, or threats.
5. Document that it is fake. A detection report showing the media is AI-generated supports platform reports, police reports, and civil claims. You can run a suspected deepfake through our detector free, 50 checks a month.
6. Talk to a lawyer about civil claims: defamation, publicity rights, and state deepfake statutes often allow damages.

Step-by-step help, including template language for reports, is in our deepfake harassment victim resources guide.

FAQ

Is it illegal to make a deepfake of someone without consent?

It depends on content and use: intimate imagery, fraud, and election deception are illegal in most of the US, while non-deceptive parody is generally protected.

Are deepfakes illegal in the US?

There is no blanket ban; deepfakes are illegal when used for specific harms such as NCII, fraud, harassment, or election deception, under federal and state law.

What is the federal law on deepfakes?

The TAKE IT DOWN Act criminalizes non-consensual intimate deepfakes and mandates platform takedowns; there is no general federal deepfake ban as of June 2026.

Are deepfake memes and parodies illegal?

Generally no. Parody and satire are protected speech, and most statutes exempt them, but the protection weakens if content is designed to deceive rather than amuse.

Can you sue someone for making a deepfake of you?

Often yes. Depending on your state, claims can include defamation, false light, right of publicity, and statutory damages under state deepfake laws.

Are deepfakes illegal in every state?

No. As of mid-2026, 47 states have deepfake laws of some kind, and coverage varies; see our deepfake laws by state table for yours.

Conclusion: Deepfake Law Is Use-Based, Not Blanket

So, are deepfakes illegal? Not as a category, and that is unlikely to change while parody and consensual uses remain protected speech. What is illegal, increasingly clearly, is using deepfakes to violate, defraud, or deceive: intimate images without consent are a federal crime, voice-clone robocalls violate the TCPA, about 30 states police election deepfakes, and fraud is fraud no matter how convincing the video.

The practical takeaways: get consent and label your synthetic media if you create, and preserve evidence and report fast if you are targeted. And when the first question is "is this even real?", you can check a suspicious video, image, or voice clip free, 50 detections a month, no card required.

This article is general information, not legal advice. Last updated June 2026; statutes cited were verified against official sources at that date.

Related reading