AI Political Ads Draw $187 Million With Few Disclosures as Platforms Struggle to Police Slopaganda
Updated
Updated · CNBC · Nov 3
AI Political Ads Draw $187 Million With Few Disclosures as Platforms Struggle to Police Slopaganda
3 articles · Updated · CNBC · Nov 3
Summary
$187 million has gone into nearly 300 AI-generated or AI-enhanced political ads this midterm cycle, with the Wesleyan Media Project saying the total is likely understated because many ads lack disclosures.
More than 90% of that spending came from Republican candidates, party committees or GOP-leaning groups, while CNBC identified undisclosed examples including a fake Kathy Hochul video from Bruce Blakeman and a Ken Paxton ad shared on YouTube and X.
Meta, Google and X all require some form of political-ad labeling or legal compliance, but there is no comprehensive federal disclosure law and roughly 30 state laws vary widely, leaving enforcement uneven.
The problem extends beyond paid ads: politicians are also spreading AI-generated organic posts, while tests by the Brennan Center found major AI models can often still be pushed to create misleading election content.
Congress has proposed bipartisan and Democratic bills to ban or expand penalties for deceptive AI campaign media, but none has advanced out of committee.
If campaigns use AI to seamlessly alter authentic speeches, will mandatory transparency labels actually change how voters emotionally react to synthetic media?
As AI deepfakes flood social feeds, can invisible watermarks truly stop bad actors from stripping labels off deceptive election videos?