Election deepfakes live in their own legal universe, and it's messier than the criminal side. The rules aren't mostly about banning the tech. They're about deception near an election. And they usually come down to disclaimers.
A growing number of states require synthetic political content to carry a clear disclaimer, especially in the weeks before an election. Some restrict knowingly deceptive deepfakes about candidates outright during that window. The details splinter fast: what counts as synthetic, how big the disclaimer has to be, who the rule applies to, what the penalty is for skipping it.
My honest opinion, having watched campaigns handle this: this is where well-meaning teams get burned most often. A fun AI-generated ad that was fine in March can be a violation in October. The calendar matters as much as the content. And 'everyone knew it was AI' is not the legal standard anywhere. The standard is whether the disclaimer was there. Smart campaign staff treat synthetic media like any regulated election material. Cleared by counsel, labeled properly, logged. The campaigns that get in trouble are the ones that treated it like a meme.
Parody and satire usually get carve-outs. But carve-outs have edges, and edges get litigated. An AI disclosure law tracker is genuinely useful here, because election-season rules change fast and the timing windows are narrow. If you're running political content with any synthetic media in it, check that state's deepfake election law and its deepfake political ad disclaimer requirements before the thing ships, not after it trends.
DeepfakeLaw tracks those disclaimer rules and election deepfake statutes across the states, with effective dates. And if you're building tools for campaigns or creators, I audit AI products for deepfake-law compliance in plain language at deepfakelaw.fyi.