Artificial intelligence has moved from the margins of American campaigning to the centre of it. Ahead of the November 3 midterm elections, political campaigns and outside groups have spent more than $137 million promoting advertisements created or enhanced with AI tools, according to the Wesleyan Media Project, a research group based at Wesleyan University that tracks election advertising with ad-tracking firm AdImpact.
The figure is a dramatic break from the last election cycle. In 2024, the same researchers identified less than $50,000 in spending on AI-generated ads. In September 2026 the tally stood at more than $80 million across 164 ads; the latest count reaches 220 AI-generated or AI-enhanced ads across 34 states.
“AI is making its way into the 2026 midterms in a big way this year,” Erika Franklin Fowler, co-director of the Wesleyan Media Project, said. “For comparison, we identified less than $50,000 in spending on AI-generated ads in the 2024 election cycle.”
What the $137 million actually counts
The headline number needs careful reading. It is the airtime spending behind ads that used AI somewhere in their production — not the cost of the AI tools themselves. The category is broad: it ranges from deepfake-style depictions of real people to cheaper uses such as generating stock-style footage or cutting production costs.
The researchers themselves warn the count is almost certainly an undercount. AI involvement is often impossible to detect unless a campaign discloses it or the imagery gives it away, and only ads the project could identify are included.
That detection problem cuts both ways for the disclosure figures. Of the ads identified, just 27% carried a disclaimer saying AI was used. That means roughly seven in ten identified ads gave viewers no such warning — but ads that disclose are also the easiest to find, so the true disclosure rate across all AI-assisted advertising may differ.
Who is spending it
Republican candidates and allied groups dominate the identified tally: 81% of the ads and 88% of the spending, against 12% of the ads from Democrats and the remainder from independent candidates or groups, according to the project.
Researchers quoted in the coverage offer one explanation for the gap: Democrats, many of whom are campaigning for stronger AI regulation, may face more reputational risk from using a tool they argue should be regulated. That explanation is an interpretation, not a finding of the spending data itself, and the spending split describes only the ads the project could identify.
The ads voters are actually seeing
Several of the identified ads do not merely polish a candidate’s image — they fabricate scenes involving real opponents.
One ad from Front Line Action, a political action committee aligned with Republican Senator Ted Cruz, targeted Michigan Senate candidate Abdul El-Sayed. It showed him playing a board game with people wearing traditional kaffiyehs, with cards reading “Death to America” and “Celebrate 9/11”, and a fabricated family photo. That ad did carry a disclaimer that it was created in whole or in part with AI. A Federal Election Commission filing cited in the coverage lists more than $92,000 in advertising opposing El-Sayed.
Another ad, produced by the campaign of Bruce Blakeman, the Republican candidate for New York governor, showed New York Mayor Zohran Mamdani and Governor Kathy Hochul strolling together and promoting a fictional “political delusion pill”. The New York Times, quoted in coverage of the research, reported that the ad included no disclaimer that it was AI-made.
In the Texas Senate race, an ad from Republican Ken Paxton’s campaign depicted Democrat James Talarico writing in a Bible with a crayon and delivering a fabricated monologue. Talarico condemned the ad as misleading and has previously tried to ban AI deepfakes through legislation.
Democratic Senator Jon Ossoff, speaking to C-SPAN, called the use of deepfake videos to fabricate political advertising “deeply troubling” and “a slippery slope toward the destruction of truth in political debate.”
A separate analysis by CNN, reported by FirstPost, found that from January through mid-September more than $16 million went to television ads that disclosed AI-generated or manipulated imagery — about 96% of it from Republican campaigns or groups — alongside at least $38 million spent on ads that appear to use AI imagery without disclosing it. That is a different measurement from the Wesleyan tally and should not be added to it.
No federal rule, a patchwork of state laws
There is no federal mandate requiring political ads to disclose AI use, though some states have legislated to restrict AI in campaign communications or require disclosures. Roughly 30 states have passed some form of law on AI in political communications, according to the coverage.
The gap matters because the technology is getting cheaper and more convincing at the same time. Campaigns can now produce elaborate, lifelike videos quickly, while voters have fewer reliable signals for telling genuine footage from fabrication. A survey of 1,568 likely 2026 general-election voters, conducted September 14–22 and cited in the coverage, found 66% were very concerned that deepfakes could change how other people vote.
AI-generated videos have already become a flashpoint in other election campaigns: French President Emmanuel Macron recently blamed AI fake videos as foreign leaders exploited France’s riots for political gain, as Emadan reported.
What happens next
With voting on November 3, the spending total is likely to rise further, and the researchers’ count will remain a partial picture until disclosure is consistent. The central question for voters is not whether AI was used in an ad — the tools are now routine — but whether the ad says so, and whether what it shows actually happened.
Until a federal disclosure standard exists, the burden falls on viewers to check the source of a striking campaign video before sharing it, and on campaigns to say plainly when a scene was generated rather than filmed.
Frequently asked questions
How much has been spent on AI political ads in the 2026 midterms?
More than $137 million, according to the Wesleyan Media Project’s latest tally of 220 AI-generated or AI-enhanced ads across 34 states, based on AdImpact spending estimates.
How does that compare with 2024?
The same researchers identified less than $50,000 in AI political ad spending in the 2024 cycle.
How many AI political ads carry a disclaimer?
27% of the ads identified by the Wesleyan Media Project included an AI disclaimer.
Is there a federal law requiring AI disclosure in political ads?
No. There is no federal mandate, though about 30 states have passed some form of law on AI in political communications.
SOURCES:
Wesleyan Media Project findings as reported by The New York Times (via Election Law Blog): https://electionlawblog.org/2026/137-million-in-a-i-generated-ads-flood-the-midterms/
AFP report via NDTV: https://www.ndtv.com/world-news/us-midterm-elections-ai-generated-political-ads-flood-us-midterms-raising-deepfake-concerns-12236548
FirstPost explainer (incl. CNN separate analysis): https://www.firstpost.com/explainers/us-midterms-2026-ai-political-ads-cross-137-million-but-who-is-regulating-them-14055463.html
