What is the deepfake law where you are?
Pick a US state or a country and read the law card the DeepfakeLaw dataset records: whether a deepfake law exists, if labeling or consent is required, what the penalties are, and whether the law is in force yet. Every field comes from the dataset; a field the dataset leaves blank is shown as not recorded, never filled in with a guess.
Loading the DeepfakeLaw dataset...
Look up a jurisdiction
Loading the DeepfakeLaw dataset...
Jump to the answer:
These cards summarize the law as recorded in the DeepfakeLaw dataset. They are not legal advice. Statutes change, regulations are pending in many jurisdictions, and court challenges can alter enforcement. Confirm with the official text linked on each card and consult qualified counsel before acting.
How to read the law card
- Status. In force means the law is enacted and currently applies. Enacted, not yet effective means it has passed but its start date is still ahead. Proposed means a bill that has advanced at least one legislative step; it is not law. Repealed means the law is no longer in force and is kept only where historically notable.
- Labeling. Shown as required only where the law imposes an affirmative duty to disclose AI-generated or manipulated content. A disclosure that exists only as a safe harbor or a defense is recorded as not required.
- Consent. Shown as required where the law requires consent to create, share, or exploit a synthetic likeness, as in non-consensual imagery and voice-likeness laws.
- Penalties. Penalty text is copied from the dataset, which takes figures only from statutory text. Where no figure was verified, the card says so instead of estimating one.
- Several laws, one place. Some jurisdictions have more than one recorded law (California has ten). Each law gets its own card, in status order, and the duties of one law are never merged into another.
- Not legal advice. This lookup is a research snapshot for information only. It must not be relied on for compliance decisions.