Researchers studying political advertising have tracked about $80 million spent on nearly 170 unique AI-generated political ads during the 2026 U.S. midterm campaigns, according to fortune.com. This election cycle marks the first time AI-generated ads regularly appear on television and social media, with Republican candidates and interest groups responsible for 83% of the spending on these ads.
The data, collected by the Wesleyan Media Project using media reports, student coders, and AdImpact, reveals a range of AI use from hyperrealistic deepfakes to subtle video enhancements. Notably, several ads feature deepfakes of prominent politicians such as Donald Trump, Nancy Pelosi, Barack Obama, Kamala Harris, and Alexandria Ocasio-Cortez. Republican advertisers have frequently used Ocasio-Cortez’s image, appearing in at least five ads. Some ads include controversial portrayals, like a Republican Senate candidate from Louisiana depicted driving a school bus full of undocumented immigrants.
The widespread use of AI in political ads raises concerns about transparency and voter deception, especially since many ads do not disclose their AI-generated nature due to inconsistent state legislation. This development highlights the growing role of generative AI in shaping political messaging and the challenges regulators face in ensuring accountability. The 2026 midterms are the first major U.S. elections to feature this level of AI-driven political advertising.
The Wesleyan Media Project has been tracking political advertising since 2010, and this cycle’s $80 million AI ad spend sets a new precedent for campaign strategies. The next major report on political ad spending is expected after the midterm elections conclude, providing further insight into AI’s impact on political communication.