AI-generated attack ads are already slopping up Mass. races
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Brendan Lynch/Axios
Deehan here with Spill of the Hill, my column unraveling Massachusetts politics.
Artificial intelligence is becoming an active political weapon in Massachusetts.
Why it matters: Two recent cases show that candidates are willing to use the technology as lawmakers struggle to regulate it.
- We're entering a world in which voters will struggle to distinguish AI-generated political content from authentic messaging.
Catch up quick: Republican gubernatorial candidate Brian Shortsleeve posted a social media ad in January using a pretty obviously artificial version of Gov. Maura Healey's voice criticizing her own record.
Yes, but: The campaign didn't disclose that it had used AI.
Separately, state Rep. Marc Lombardo posted an AI-generated image in a campaign attack against an opponent running for his Billerica House seat.
- That fake image posted to Facebook was a mock newspaper article with challenger Daniel Darris-O'Connor joining hands with New York City Mayor Zohran Mamdani at the sort of physically impossible angle AI image generators are known for.
- The ad meant to link Darris-O'Connor to Mamdani's democratic socialist policies.
- Lombardo did not respond to Axios' request for him to explain the ad.
Between the lines: Both examples exist in the legal gray zone of AI political ads.
- Shortsleeve's campaign called its synthetic-voice ad parody. Lombardo's AI image reads like opposition research factoids, even though it's made to look like an old-timey newspaper article.
- Current law largely turns on whether political content is "materially deceptive" to an audience. It's a standard that's easy to dispute.
The big picture: The barrier to entry into AI art and video is long gone.
- A realistic fake audio clip or AI-generated attack image no longer takes campaign resources or know-how. It's just a few clicks away for any state-wide or down-ballot campaign aide.
What's next: State lawmakers are considering disclosure requirements and limits on deceptive AI campaign content. But in the meantime, the 2026 election cycle may turn into the wild west of AI ads.