Tracking uncovers $80 million spent on AI political ads

The 2026 U.S. midterm campaigns are officially the first election cycle in which artificial intelligence-generated political advertisements are regularly appearing on voters' televisions and social media feeds. As researchers studying political advertising through the Wesleyan Media Project since 2010, our team has closely monitored this evolving landscape. Utilizing a combination of data from media reports, diligent student coders, and AdImpact—a prominent firm that tracks political ad spending—we have tracked about $80 million in spending on almost 170 unique ads that utilize AI.

This milestone highlights a massive technological shift in how political campaigns attempt to persuade voters. Generative AI tools have made it cheaper, faster, and easier than ever to create hyper-realistic imagery, synthetic voiceovers, and entirely fabricated video scenarios. Candidates and political action committees are increasingly leveraging these capabilities to produce compelling content at a scale that was previously impossible. However, the rapid adoption of this technology has significantly outpaced regulatory frameworks and platform policies.

One of the most alarming findings from our tracking is that the vast majority of these AI-generated ads are being deployed without any clear disclosure. Viewers watching television or scrolling through social media feeds are often left unaware that the figures, environments, or statements they are witnessing have been synthetically manufactured. This absence of labeling deprives voters of critical context, making it extraordinarily difficult for the average citizen to distinguish between authentic documentation and sophisticated digital fabrication.

The implications of undisclosed AI ads on democratic processes are profound. In an era where trust in public institutions and media is already fragile, the proliferation of deceptive synthetic content threatens to further distort public discourse. Voters may be swayed by fabricated controversies or idealized candidate portrayals that have no basis in reality. Furthermore, because different social media platforms enforce varying rules regarding synthetic media disclosures, the regulatory landscape remains fragmented and largely ineffective.

As the 2026 midterms progress, the findings from the Wesleyan Media Project underscore an urgent need for comprehensive transparency rules. Lawmakers, tech platforms, and election administrators must work together to establish clear standards for labeling AI-generated political content. Without meaningful oversight and mandatory disclosures, the integrity of future elections will remain vulnerable to the unchecked influence of deceptive media technologies.