AI Regulation Tracker / Enacted law
Louisiana Now Requires AI Disclosure in Political Ads, With Jail Time on the Table
Louisiana enacted Act 806 (House Bill 459), a law that joins the wave of state election-AI/deepfake disclosure laws requiring a clear disclosure when artificial intelligence is used to create or manipulate a candidate's image, likeness, or voice in electioneering communications. It takes effect August 1, 2026, and a knowing violation carries a criminal penalty of up to two years. This is a binding statute, not guidance.
On July 13, 2026, Louisiana joined the wave of states putting a criminal disclosure rule around AI-generated content in political campaigns. House Bill 459, carried by New Orleans Representative Mandie Landry, was enrolled as Act 806 and signed into law by Governor Jeff Landry. It takes effect August 1, 2026. The law does not ban AI in campaign material. It requires that you tell people when the candidate they are looking at or listening to was built or altered by a machine.
What the law actually requires
Act 806 amends the Louisiana Election Code and adds new provisions at R.S. 18:1463(H) and (I), with a supporting definition added at R.S. 18:1483(33). The core of it is a disclosure duty tied to electioneering communications. When a communication portrays a candidate, or an elected official subject to recall, and that portrayal was created, altered, or digitally manipulated using artificial intelligence, the communication has to carry a clear and understandable disclosure that AI was used.
The trigger is not every touched-up photo. The statute reaches AI-generated depictions that would, in its words, "falsely appear to a reasonable observer to be an authentic record of the actual speech or conduct of a candidate or an elected official subject to recall," or that would falsely swap someone else's likeness for the candidate's. That is the deepfake case: a fabricated video or audio that a normal viewer would take as a real record of the person actually saying or doing something. Routine editing that no reasonable person would mistake for an authentic record is not the target.
The disclosure itself has to be clear and understandable and prominently placed on the communication, in line with how Louisiana already handles other required campaign disclosures like third-party payor notices. In other words, this is not a buried footnote standard. The point is that a voter can tell, at a glance, that what they are seeing was made or changed by AI.
The penalty is criminal, and that is the part to sit with
This is where Act 806 separates itself from a lot of the AI-labeling conversation. A violation is not just a civil fine or a slap from an ethics board. Done with the required intent, failing to disclose is a crime. The penalty is a fine of up to ten thousand dollars, imprisonment for up to two years with or without hard labor, or both.
The bill's author put the stakes in plain terms. As Representative Landry described it in reporting on the law, "If you are able to prove that I did an AI video of you robbing a bank when you were running for office, and I didn't have a disclosure, I could be fined and face up to two years in jail." That is the scenario the statute is built for: a convincing fake, aimed at a candidate, pushed out during a race, with no label telling voters it was synthetic.
Because the serious penalty attaches to intentional conduct, this is not a trap for an honest mistake in the abstract. But intent in an election-fraud context is exactly the kind of thing prosecutors and opposing campaigns litigate hard. If your shop is producing AI-assisted creative, "we didn't mean it that way" is a defense you argue after you have already been charged. The cheaper path is the disclosure.
Why this one matters beyond Louisiana
Two reasons. First, it is a US state drawing a bright line that AI depictions of candidates are a distinct category of campaign speech with their own disclosure rule and their own criminal exposure. Federal law here is patchy, and states are the ones actually writing enforceable rules. A firm that runs national or regional campaigns cannot treat Louisiana as a carve-out. The safer operating assumption is that AI-in-politics disclosure rules are now a live, state-by-state compliance surface that will keep growing, and Louisiana is a concrete data point on where the line and the penalty are being set.
Second, the effective date is close. August 1, 2026 is not a distant phase-in. Any campaign, consultant, or agency touching a Louisiana race after that date is operating under this rule immediately. If AI is anywhere in your production pipeline, the disclosure question is now a pre-publication checklist item, not a next-cycle project.
What campaigns, consultants, and counsel should do now
Inventory where AI touches candidate depictions. Any tool that generates, alters, or manipulates a candidate's face, body, or voice is in scope for the disclosure analysis. That includes video, audio, and image work, and it includes third-party creative you place but did not make. Build a disclosure default. If a piece uses AI on a candidate's likeness in a way a reasonable viewer could read as authentic, attach the clear disclosure by default rather than deciding case by case under deadline. Treat the reasonable-observer test as the line. The legal exposure turns on whether the depiction could pass as a genuine record of what the person said or did, so that judgment call is the one to document. And get election-law counsel on the intent question before you rely on it. The reduced-risk position is disclosure, not betting that a prosecutor will not be able to prove intent.
None of this requires abandoning AI in political creative. It requires labeling the depictions that could fool a voter, and doing it prominently, before the piece runs in Louisiana.
Questions professionals are asking
Does Louisiana Act 806 ban using AI in political ads?
No. It does not prohibit AI in campaign material. It requires a clear and understandable disclosure when AI is used to create, alter, or digitally manipulate a candidate's image, likeness, or voice in an electioneering communication that would falsely appear to a reasonable observer to be an authentic record of the person's speech or conduct.
When does the law take effect?
August 1, 2026. The bill, House Bill 459, was enrolled as Act No. 806 and signed by Governor Jeff Landry. Any communication in a Louisiana race after that date is subject to the disclosure rule.
What is the penalty for not disclosing?
A violation committed with the required intent is a crime. The penalty is a fine of up to ten thousand dollars, imprisonment for up to two years with or without hard labor, or both. The serious penalty attaches to knowing conduct rather than an innocent mistake, but intent in an election context is heavily litigated.
Does every edited campaign photo need a disclosure?
No. The statute targets AI-created or AI-manipulated depictions that could falsely appear to be an authentic record of what a candidate actually said or did, or that swap someone else's likeness for the candidate's. Ordinary editing that no reasonable observer would mistake for an authentic record is not the focus. The reasonable-observer test is the line to document.
Does this affect campaigns based outside Louisiana?
It applies to electioneering communications in Louisiana races, so out-of-state campaigns, consultants, agencies, and PACs are covered when their work touches a Louisiana election. It is also an early US state model, so firms running multi-state programs should expect similar AI-disclosure rules to keep appearing and treat this as a live compliance surface.
RELATED BRIEFINGS
Browse the full AI Regulation News tracker
Informational analysis for working professionals, not legal advice. Confirm how this statute applies to a specific communication or campaign with qualified election-law counsel licensed in Louisiana.