Talarico Accuses Paxton of Using AI Ad to Fabricate His Words in Texas Race

Texas Democrat James Talarico is accusing Republican Attorney General Ken Paxton of releasing a campaign ad built with artificial intelligence that fabricates statements in Talarico's voice. In a public response, Talarico said the ad amounts to Paxton "lying to the people of Texas" by putting "words in my mouth" that he never spoke. The dispute marks one of the most direct confrontations yet between a candidate and an opponent over synthetic media in a statewide race.

The ad in question uses AI voice cloning or manipulation to make it appear Talarico said things he did not say, according to his account. Talarico, a state representative and seminary graduate who has built a following around education and healthcare issues, did not dispute the general subject matter of the ad so much as its method: attributing invented language to a real candidate. His pushback came as searches for his name spiked past 500, signaling the exchange traveled well beyond Texas political circles.

Paxton, who has served as attorney general since 2015 and has been a frequent national figure on conservative legal causes, has not backed away from the spot. The two are on a collision course in a state where AI-generated political content sits in a gray zone: federal rules do not specifically ban deceptive synthetic campaign ads, and Texas has no statute that squarely addresses voice cloning in elections. That leaves candidates to fight such claims in the press and before voters rather than in a regulatory forum.

The episode fits a broader pattern from the 2024 and 2025 cycles, when campaigns and outside groups tested AI tools for fundraising appeals, rapid-response videos and negative mail. Most of those uses were disclosed or obvious. What Talarico describes is different: a spot that presents fabricated speech as real, which election law scholars say is the hardest category to police without running into First Amendment limits. Platforms including Meta and Google have rolled out labeling requirements for AI political content, but enforcement depends on detection and voluntary compliance.

The immediate stakes are local, but the precedent is not. If a statewide ad can put invented words in a rival's mouth without legal consequence, other campaigns will treat that as permission rather than a red line. Watch whether Paxton's team confirms how the ad was produced, whether Talarico files a complaint with state election authorities, and whether Texas lawmakers take up synthetic-media disclosure rules when they next convene. The open question is whether voters punish the tactic or simply absorb it as another feature of modern campaigning.