A wave of state laws built to stop AI-generated fake political ads and images is about to face its first real-world stress test. With early voting for the 2026 midterms approaching, 31 states now have laws on the books specifically targeting election-related deepfakes, according to a tracker maintained by the watchdog group Public Citizen — up sharply from just a handful a few years ago.

The laws vary widely in what they actually do. Most require disclosure or disclaimers when AI has been used to generate political content, and many give candidates targeted by non-compliant synthetic media legal recourse in court. Some states empower election officials to demand corrections or takedowns, while others impose direct penalties on anyone who knowingly distributes deceptive AI-generated content close to an election.

A Patchwork With Real Gaps

A few states have gone further than most. Minnesota and Texas both prohibit distributing deepfakes within a set window of days before an election, while Maryland's ban applies year-round rather than just during campaign season, the Washington Examiner reported. But the legal foundation underneath all of it remains shaky. Courts have already struck down two state election-deepfake laws on First Amendment grounds, and legal experts say more challenges are likely as campaigns increasingly lean on generative AI tools to produce ads, robocalls and viral images.

COMPANIES OPERATING ACROSS STATE LINES SHOULD EXPECT "CONTINUED VOLATILITY" THROUGH NOVEMBER.

That's the warning from legal analysts tracking the patchwork, who note that a synthetic ad perfectly legal in one state could trigger civil liability — or a First Amendment fight — the moment it crosses into a neighboring one. Congress, meanwhile, still hasn't passed any federal deepfake legislation of its own, leaving the state-by-state approach as the only real backstop heading into November.

Already Showing Up in Campaigns

The stakes aren't hypothetical. Reporting from the Arizona Capitol Times found AI-generated content already surfacing in campaign material ahead of primaries this summer, as synthetic video and voice-cloning tools have become cheap and widely accessible. Election officials and campaign lawyers in multiple states say they're bracing for a surge of AI-manipulated content in the final stretch before Election Day, particularly in competitive House and Senate races where a single viral fake could scramble a close contest.

NBC News reported the new AI-election rules are among a broader crop of state laws taking effect around the midterms, alongside new paid-leave mandates and changes tied to rising Affordable Care Act premiums. Pluribus News, which tracks state policy, called the coming months a genuine test of whether these laws can actually hold up once real money and real campaigns start pushing against them.

With primaries already underway in several states and the general election just under three months out, the coming weeks will show whether the current patchwork of rules is enough — or whether it takes a high-profile deepfake scandal, and likely another lawsuit, to find out where the real limits are.

Enforcement is also an open question. Several of the state laws rely on candidates or watchdogs to spot manipulated content and file complaints after the fact, rather than giving officials the tools to catch synthetic media before it spreads — a gap that becomes more consequential the closer a fake gets released to Election Day. Platform companies, for their part, have been left to interpret a different rulebook in nearly every state where they run political ads, adding another layer of compliance headaches on top of the legal uncertainty.