In brief
AI‑driven disinformation is surging on social media, with 220 ads valued at over $137 million and no disclosure of AI use. While only 1 % of voters see AI as a credible source, a fifth plan to consult AI for ballot research, and a cottage industry advises candidates on phrasing. Tech firms lobby against tighter AI regulation, further complicating accountability.
AI‑driven disinformation has rolled into the 2026 U.S. midterm elections, flooding social media feeds with political ads that conceal their artificial origin.
AI Advertising Surge
The Wesleyan Media Project – a nonprofit that tracks AI‑generated political advertising – identified 220 such ads on October 3, carrying an estimated value of more than $137 million. That figure stands in stark contrast to the April 2024 cycle, when total spend reached only about $50 000. The projection that this initial estimate is probably low reflects the difficulty of detecting AI involvement.
"AI is making its way into the 2026 midterms in a big way this year," the Project said in a statement, highlighting how cheaply and rapidly disinformation can be produced. The absence of disclosure, according to analysts, risks misleading voters and eroding public trust in elections.
Frances Haugen, a former Facebook data engineer turned whistleblower, warned that AI “creates an information operation that is super cheap and as high quality as previously impossible.” She added that content moderation loopholes can be bypassed by “tweaking the post a little bit each time you create them,” and some ads will slip through the cracks.
Some of the most provocative examples include a clip of a Republican gubernatorial hopeful hugging former President Clinton with the caption “Clinton Lover, Trump Hater.” Another shows Democratic Senate candidate Abdul El‑Sayed laughing at cards reading “Death to America” and “Celebrate 9/11.” A third depicts Republican primary candidate James Fishback hosting a party with Iranian leaders wielding scimitars at the Star of David, a symbolic target of Judaism.
Because most of these pieces carry no disclosure that AI has generated them, their believability can distort public perception and weaken the institutional integrity of the electoral process.
Broader Misinfo Landscape
A recent survey by the American Association of Political Consultants Foundation found that only one percent of respondents regarded AI as their most credible source of election information, yet 20 % plan to use AI to research a race on their ballot. The same survey reported that half that proportion had already received election details through an AI‑generated summary, indicating a growing reliance on the technology even among wary voters.
The surge in AI content has also sparked the emergence of a private consulting sector, with the left‑wing advocacy group Run for Something launching CampSight, a platform designed to track political ads and help candidates shape AI‑written narratives. This cottage industry signals how political campaigns are increasingly outsourcing messaging strategies to algorithms, raising questions about representation and authorial accountability.
Tech firms have begun lobbying against stricter AI regulation, a move that could further obfuscate accountability and limit efforts to curb the spread of misinformation. Critics argue that without disclosure requirements, voters cannot reliably assess the authenticity of campaign messages.
Additional concerns come from AI chatbots, which, according to a recent report from The Star, are disseminating misleading voter information, misidentifying candidates, and repeating Russian propaganda. These chatbots add another layer to the complex information environment that voters must navigate before casting their ballots.
Across the board, the convergence of undisclosed AI‑generated ads, a burgeoning consulting industry, and lax regulation threatens the core civic stakes of federal elections: institutional integrity, transparent dialogue, and informed public participation. In the weeks before November 3, the electorate faces a battleground not only for positions of office but for the very mechanisms that certify the fairness of democracy itself.
