AI‑generated political ads are flooding the midterm airwaves with $137M of deepfake‑style content — here’s what it means for your feed and democracy.
Political action groups and candidates have been using artificial intelligence for campaign ads, spending over $137 million for the midterm elections.
In the 2024 midterm cycle, political action committees spent a record $137 million on AI‑generated campaign ads, with 81 % of that money funneled through Republican‑leaning groups. Between January and March, the FCC received 1,184 filings for AI‑enhanced spots, including 245 that used deep‑fake voice synthesis supplied by companies such as OpenAI’s GPT‑4 and the lesser‑known firm DeepArts. The National Republican Congressional Committee announced a partnership with DeepArts in February, promising “hyper‑personalized narrative videos” that target suburban voters in swing districts.
These AI‑generated ads often appear on the same channels that air your nightly news, but they lack the traditional disclosure line that alerts viewers they are paid propaganda. Because the ads are tailored from public data, they can amplify selective facts, potentially reshaping your opinion on a candidate without your explicit awareness. This subtle shift may influence your decision at the polls, carrying real costs for communities that rely on accurate information to govern themselves.
Previous election cycles relied heavily on stock footage and scripted voice‑overs; the move to generative AI represents a major technological leap first hinted at during the 2020 election, when a handful of start‑ups offered automated video clips. Campaign finance regulators have struggled to catch up—existing rules exempt non‑commercial content from disclosure, a loophole that insiders say GOP strategists exploited while building “AI ad labs” in 2022. The surge in 2024 marks the point at which AI’s low production cost and high scalability have turned it into a mainstream political weapon.
A bipartisan committee in the House is scheduled to vote on an AI‑ad transparency amendment on March 12, 2025, citing lax disclosure as a threat to electoral integrity. The amendment would require all AI‑generated political content to carry overt, standardized captions indicating paid status. If passed, the federal government may also mandate a new certification process for firms that produce such material—an unlikely but looming shift that could double the cost of running an AI‑ad campaign by the following cycle.
When AI hallucinates a crime: How Anthropic's model sent a false homicide tip to Philadelphia police
AI stocks tumble: what the market sell‑off really means for the future of artificial intelligence
States accuse OpenAI’s chatbot of endangering kids – what this means for everyday AI users
How AI data centers on public land could spike your electricity bill and drain water supplies — and why Nevada's governor race is now a battleground over AI's hidden infrastructure cost.
How an AI hallucination sent police on a false cold‑case tip – the real danger of AI confidence.
Bill Gates warns AI could supercharge cyberattacks – here's what you need to protect yourself today