Raleigh-Durham/ Politics & Govt

The Wilmington AI Ad Fight Is Testing What Voters Can See, Verify and Challenge

AI Assisted Icon
Published on September 16, 2026
The Wilmington AI Ad Fight Is Testing What Voters Can See, Verify and ChallengeW. Jones St. — Downtown Raleigh Street Scene
Google Street View

An AI-generated campaign ad attacking North Carolina Senate candidate Jessica Bichler has turned a Wilmington race into a test of consent, accuracy and disclosure. The commercial, paid for by Republican Sen. Michael Lee’s campaign, depicts Bichler meditating and contorting her body while associating her with positions including defunding police and providing free health care to undocumented immigrants, according to WWAY News 3.

Bichler, who co-owns the Wilmington Yoga Center, says the campaign used her face and body without permission. She sent Lee’s campaign a cease-and-desist letter, but the central dispute was not simply whether artificial intelligence was used. Reporting cited by NPR says her letter challenged the way the ad portrayed her policy positions. Bichler has said AI is likely to remain part of political advertising, while ruling out its use in her own campaign.

The fight is over more than a disclaimer

Lee’s campaign has defended the commercial as obvious comedy or parody and said it voluntarily included disclosures that AI was used. The campaign later released a revised version with toned-down claims and a substantially larger disclaimer, NPR reported. The dispute leaves several questions unresolved: whether viewers understood the imagery as fictional, whether the policy claims were accurate, and what legal remedy—if any—would apply to a political message that uses synthetic imagery without claiming the underlying footage is real.

The ad’s reach also illustrates why labeling matters. WECT reported that the original video had more than 550,000 YouTube views by an afternoon in early September, while some Bichler supporters reportedly believed the AI-generated yoga footage was authentic. A view count does not show how many people were persuaded, and the reporting does not establish that the ad violated a law.

North Carolina’s policy question is still unsettled

North Carolina lawmakers have considered disclosure requirements, but the relevant measure remains a proposal rather than enacted law. The UNC School of Government’s Legislative Reporting Service says H 375 would require a clear and conspicuous disclosure for certain paid political communications containing synthetic media. That legislative summary does not establish that the proposal governed the Lee-Bichler ad or that it became law.

The distinction is important because disclosure rules and truth-in-advertising disputes address different problems. A label can tell viewers that an image was synthetically altered; it does not by itself establish whether a candidate actually holds the views attributed to her. In the Wilmington case, the cease-and-desist letter and the revised commercial put the accuracy of those representations at the center of the conflict.

A broader trend, with limits to the comparison

The Wesleyan Media Project counted at least 164 political advertisements nationwide that had been produced or enhanced by AI during 2026, according to WUNC. The figure measures use of the technology, not deception, illegality or electoral impact, and it does not show how many ads appeared in local races.

North Carolina had an earlier, separate example during the 2024 gubernatorial campaign, when political independent Todd Stiefel created an AI-generated ad depicting then-Republican candidate Mark Robinson with 14 fingers, WUNC reported. That ad is a historical precedent for synthetic political imagery in the state, but it involved a different race, sponsor and election; it cannot establish what happened in the Bichler-Lee contest.

The policy landscape is also being shaped by the platforms carrying these messages. The Associated Press reported that Google requires political ads using AI to synthetically alter imagery or sounds depicting people or events to carry a prominent disclosure on Google and YouTube. Meta’s published political-ad policy covers advertisements that depict a real person saying or doing something they did not, or a realistic-looking event that did not happen. Platform rules may affect distribution, but they are separate from North Carolina’s proposed legislation and do not resolve the candidates’ dispute over the ad’s claims.

For voters, the Wilmington controversy offers a practical warning without proving that every AI political ad is deceptive: treat synthetic imagery as a campaign argument, not as evidence that an event occurred. The remaining questions in this race concern the ad’s factual claims, the effect of its revised disclosure and whether Bichler or the state can obtain a remedy beyond a campaign’s voluntary changes.