Social Rocket
In the US midterms 2026, campaigns and political groups have spent more than 187 million US dollars on ads featuring AI-generated or AI-edited content. The Wesleyan Media Project counts 29 distinct ads in its analysis from October 289. The finding matters to advertisers beyond politics too: artificial intelligence has changed how campaign material is produced, while transparency about its use lags behind.
Within two weeks, roughly 50 million dollars more came in
In the previous analysis from October 15, the researchers had recorded 220 AI-assisted ads with more than 137 million dollars in ad spend. Two weeks later there were 69 additional creatives and at least 50 million dollars more. For comparison, for the entire US election cycle 2024 the project had identified fewer than 50.000 dollars for such spots.
The tally covers ads in which AI fully generated content or edited individual elements, from synthetic images to manipulation of existing material. The millions cited are spending on distributing these ads, not the cost of AI production.
90 percent of spending came from the Republican camp
According to the latest Wesleyan analysis, 169 million of the 187 million dollars, or a good 90 percent, come from Republican candidates, party committees or affiliated groups. They are responsible for 82 percent of the identified creatives; Democratic actors account for 13 percent. The data was combined from AdImpact ad spending, labels, reporting and review of ad creatives.
The gap between creative count and spending is revealing: for reach, what matters is not only how many spots a campaign produces, but how intensively it runs them. The overall US election ad budget of just under 11,9 billion dollars and the shift to connected TV show the environment in which these AI campaigns competed.
Labeling remains the weak point
In the October interim analysis, only 27 percent of the 220 AI ads recorded at the time carried a notice of AI use. The later publication contains no correspondingly updated figure. The Wesleyan team also assumes that inconspicuous AI edits are missing from its survey.
For brands the point is practical: a technically perfect spot can be spread on social media in seconds; its origin and editing should be just as quickly recognizable. Especially with realistic-sounding voices, faces and statements, a clear approval and labeling routine matters more than the next production shortcut.
Different rules apply in Germany
In the EU, the 10 Regulation on the transparency and targeting of political advertising2025 has been fully applicable since October . Political ads must, among other things, disclose the sponsor, spending and, where applicable, targeting information. On 12 October 2026, the EU Commission also launched a pilot version of the European online ad repository. The required political advertising disclosures and a notice on AI-generated content serve different transparency purposes.
For German social media teams there is also the fact that Meta has stopped political advertising in the EU. Anyone spreading political messages via creators, other platforms or their own channels must therefore clearly separate the specific ad format and its labeling obligations. How difficult this gets with paid influencer posts is shown by our look at the political creator campaigns of the midterms.
The key marketing lesson is not to forgo AI creatives, but to plan production, approval, documentation and delivery together. A faster creative process only gains value when teams can at the same time track which visual runs in which version on which channel. TikTok's election safety measures with AI labels also show how closely creative questions are now tied to platform rules.
Sources: Wesleyan Media Project, 29 October 2026 · Wesleyan Media Project, 15 October 2026 · EU Commission · ADWEEK commentary from 5 November