How James Talarico’s opponents use AI to boost false claims - PolitiFact

Texas U.S. Senate candidate James Talarico has become the target of a series of AI-generated political advertisements that depict him in scenarios he never actually experienced. These digital fabrications are part of a growing trend in political campaigning where artificial intelligence is used to portray opponents saying or doing things they never did.

What happened

The nonprofit group Citizens for Sanity published several AI-generated ads mocking Talarico, including one where he appears to wear a dress while singing a parody of a song from "The Sound of Music." Another ad portrays him as a disheveled preacher claiming "God is nonbinary." While these videos present a lifelike version of the candidate, the events depicted are entirely artificial.

Beyond these specific videos, the campaign has faced other uses of synthetic media. In March, the National Republican Senatorial Committee released an ad featuring an AI-generated Talarico reading his own real social media posts, and in June, Texas Governor Greg Abbott posted an AI-generated image depicting a female version of the candidate. As of June 22, the ads from Citizens for Sanity had garnered roughly 1.4 million impressions on Facebook.

While Texas law prohibits political deepfakes within 30 days of an election, these incidents are occurring months ahead of the midterms, highlighting a gap in regulation. Experts note that this shift signifies a departure from the 2024 election cycle, where campaigns were generally hesitant to use generative AI. Observers warn that such tactics force candidates to spend valuable time debunking false narratives, potentially making it increasingly difficult for voters to distinguish between truth and fiction.