AI Video, Consent, and Your Rights: What the Law Protects

Consent is the line that runs through this entire topic. If a video depicts a real, identifiable person in a sexual context, that person has to have agreed to it — and if they didn't, the content is harmful and, increasingly, unlawful, no matter how it was made. Your face and body are yours to control; the right to say no to being depicted this way is exactly what the newest laws are built to protect.

Fictional characters sit on the other side of that line. An invented person is no one, so there is no one to consent and no one to harm. That is why fictional-by-default is both the ethical lane and the legal one: the creativity happens in a made-up world, and no real person's rights are in play unless that person has explicitly agreed.

In May 2025 the United States enacted the TAKE IT DOWN Act, its first federal law aimed squarely at non-consensual intimate imagery. It makes it a federal crime to knowingly publish intimate visual depictions of an identifiable person without consent — and it explicitly covers AI-generated "deepfakes," closing the old "it's not really them" loophole. Sexual depictions of minors are covered with no exceptions.

The law also puts a duty on platforms. Covered online services must remove reported non-consensual images within 48 hours of a valid request and make reasonable efforts to take down identical copies, with the Federal Trade Commission enforcing that takedown obligation. In practice, that gives victims a fast, concrete route to removal rather than a years-long fight.

Where the TAKE IT DOWN Act adds criminal and takedown pressure, the DEFIANCE Act is designed to give survivors a civil right of action — the ability to sue the people who create, distribute, or knowingly hold sexual "digital forgeries" of them, and to recover substantial damages. It targets the AI-fabricated sexual deepfake specifically, treating the victim's lack of consent as the core wrong.

This one is still moving through Congress: the U.S. Senate has passed it and it has been reintroduced, so treat it as the clear direction of federal law rather than a settled statute. Many U.S. states already provide their own civil and criminal remedies for sexual deepfakes today, so a victim is rarely without options even before the federal civil right lands.

Two categories are illegal almost everywhere and carry serious penalties: any sexual depiction of a minor — real or fully AI-generated — and non-consensual sexual imagery of a real, identifiable adult, including deepfakes. "It's only AI" is not a defense for either. If the output sexualizes a real person without consent, or depicts anyone who is or appears underage, it is unlawful regardless of how it was produced.

Fully fictional, adult-only content is treated very differently. In much of the world it is handled like adult fiction in any other medium — a novel, a comic, an animation — because no real person's likeness or consent is involved. The dividing line is not "is it explicit" but "is a real, non-consenting person in it."

Other regions are converging on the same principle from different angles. The European Union's AI Act adds a transparency duty: AI-generated or manipulated media, including deepfakes, must be disclosed and labeled as artificially generated. The United Kingdom's Online Safety Act 2023 requires adult platforms to use highly effective age assurance and, alongside long-standing obscenity rules, can restrict even purely fictional sexual material.

Rules still vary a lot by country and are changing quickly. Japan permits fictional adult content but requires mosaic censorship of explicit detail; some countries apply obscenity or "non-photographic image" laws to fiction as well. The throughline everywhere is consent and age — real people need to agree, and no one underage is ever permitted — so check the rules where you actually live.

Act quickly and keep evidence. Report the content to the platform hosting it and cite non-consensual intimate imagery; in the U.S. the TAKE IT DOWN Act gives covered services 48 hours to remove it once you send a valid notice. For adults, StopNCII.org can hash your images so participating platforms block re-uploads; for anyone under 18, report to NCMEC and its Take It Down service. Preserve links, screenshots, and dates in case you pursue a claim.

You do not have to handle it alone, and you should never retaliate by generating content of your own. If the misuse involves a platform you use, use its abuse-reporting path — on Nureta, that means its content policy and report route. The whole point of a consent-first design, with no face-swap and no real person without their consent, is that this category of harm should never originate there in the first place.