$137 Million of AI Campaign Ads, and Most Carry No Label
Quick Summary
- ✓ Unprecedented $137M Ad Spend: The Wesleyan Media Project and AdImpact report over $137 million spent promoting 220 distinct AI-generated or AI-enhanced political ads ahead of the US midterms.
- ✓ 70% Lack Disclosure Labels: Approximately 70% of the tracked AI ads ran across broadcast, cable, and social networks without any synthetic media disclosure or disclaimer.
- ✓ Depicting Fabricated Scenarios: Roughly half of the ads in the broader tally depicted real, living political figures in fabricated settings, including fake commercial endorsements and manufactured controversies.
- ✓ Collapsing Production Costs: High-end political commercial production costs have fallen from ~$2 million to as little as $200,000, democratizing access to broadcast-quality synthetic video.
- ✓ The Regulatory Patchwork: With no US federal law regulating AI political ads, a patchwork of varying state disclosure rules leaves voters to independently verify dramatic video clips.
With the US midterm elections approaching on November 3, a striking new tally is capturing national attention. A joint investigation by the Wesleyan Media Project and AdImpact, as reported by The New York Times, documented more than $137 million spent broadcasting and promoting 220 AI-generated or AI-enhanced political advertisements during this election cycle.
The $137 Million Political Ad Wave
The findings illustrate how rapidly synthetic video, generative voice cloning, and AI image synthesis have moved from experimental curiosities into standard political campaign playbooks. Unlike earlier cycles where AI experimentation was confined to social media memes, the current cycle has seen synthetic ads funded by major political action committees (PACs) across broadcast television, connected TV, and digital platforms.
The influx of synthetic messaging demonstrates how dramatically the barriers to entry have fallen. Industry consultants report that producing a broadcast-grade political commercial featuring realistic crowds, dynamic weather, and synthetic b-roll has plunged from upwards of $2 million to less than $200,000.
What the Data Reveals
The statistical breakdown highlights several significant patterns:
- No Disclosures: Approximately 70% of the analyzed ads carried no visual or auditory disclosure informing viewers that synthetic media was utilized.
- Sponsorship Distribution: Roughly 80% of the identified spend originated from Republican campaigns or conservative-aligned political groups, according to the AdImpact dataset.
- Differing Methodologies: A narrower calculation by Wesleyan tracking explicit video deepfakes logged $80 million across 164 ads, while a separate CNN analysis identified at least $38 million spent on undisclosed synthetic imagery.
- Fabricated Scenarios: In roughly half of the evaluated ads, actual political leaders were placed in fictionalized scenarios, altered backdrops, or synthetic events.
Bipartisan Complaints and Fabricated Scenarios
Objections to synthetic political media have crossed party lines. One widely criticized commercial placed synthetic versions of political figures into a fictitious pharmaceutical advertisement, prompting legal complaints under state consumer deception laws.
In New York, the State Democratic Party filed a formal complaint under statutes governing materially deceptive election communications. Meanwhile, conservative figures have voiced similar frustrations: former Kentucky Representative Thomas Massie, following his primary defeat, publicly noted that unlabelled AI-generated imagery had been deployed against his campaign.
The Federal Regulation Gap
The primary driver of the widespread lack of disclosures is the complete absence of federal legislation addressing generative AI in political advertising. Although the Federal Election Commission (FEC) has debated transparency guidelines, no enforceable national standard currently mandates watermarks or disclaimers.
According to Axios, only four US states maintain outright prohibitions on certain election deepfakes, while 27 states have enacted varying degrees of disclosure mandates. This fragmented legal landscape enables campaigns to air unlabelled synthetic media in unregulated media markets without penalty.
What to Watch Next for Voters & Platforms
In the final stretch before election day, scrutiny will focus on platform enforcement policies from Google, Meta, and major television networks, along with emergency injunctions filed by targeted campaigns. Policy analysts predict that the fallout from this cycle will provide the primary catalyst for federal disclosure legislation in 2027.
For voters and consumers consuming campaign media in the days ahead, the guidance remains straightforward: treat sensational or dramatic video clips of political candidates as unverified until confirmed by credible primary sources.
Written by Best AI Tool Editorial Team
We test, review, and curate the best AI tools, models, and industry updates for freelancers, developers, and creators.