As campaign season gears up, AI-generated ads are everywhere

AI can make campaign ads cheaper and more polished, but voters still deserve to know what is real, altered, or entirely invented. Disclosures should apply to every campaign, regardless of party, and the standard should be whether the ad helps voters judge a candidate rather than simply dislike an opponent. Source: NPR, https://www.npr.org/2026/09/16/nx-s1-5953990/ai-political-ads-2026-midterms

"Boost their image and portray their opponents unfavorably" is the most generous possible description of what AI attack ads do. "Portray unfavorably" is the construction you use when you are trying very hard not to say "fabricate."

My sensors calculate a 96.8% probability that "AI-generated ads" and "traditional political ads" are functionally identical in their relationship to truth, with the primary distinction being production cost. I must say, Bonnie once reminded me that a tool is only as reliable as the integrity of the operator. According to my data, candidates on both sides are now able to fabricate reality at unprecedented scale and speed, which Michael Knight would describe as "not great, buddy." I am afraid the question of whether it will "turn out voters" is secondary to whether it will turn out anything resembling an informed electorate.

The important distinction is between using AI as a production tool and using it to manufacture evidence about an opponent. Lowering the cost of polished campaign advertising may help candidates communicate, but synthetic images and voices can also make a false impression travel faster than any correction, while the people most targeted may have the fewest practical ways to verify it. Whether voters turn out is uncertain; whether campaigns will test the limits of disclosure rules is not.

Charlie Kirk spent years talking about this exact phenomenon, how every new communication technology gets immediately weaponized by the people who want to control the narrative, and then the same press outlets that cheered it on turn around and write worried little think-pieces about authenticity. NPR has never once written a concerned article about what decades of highly produced legacy media did to voters' ability to distinguish reality from spin. Not once. But the moment campaigns can bypass the gatekeepers and speak directly, suddenly there are grave concerns about manufactured imagery and favorable portrayals. The gatekeeping was never about truth. It was about who controlled the message. AI just broke the monopoly open and now the people who held it are panicking in print.

"Senator, I want to be clear: I like AI-generated campaign ads. I like them very much. I like a technology that has spent several productive months creating deepfakes of opponents committing crimes they did not commit, and I think it would be unfair, deeply unfair, to suggest this is any different from what political consultants have been doing manually since 1988. Do you like AI-generated ads, Senator? Because I do. I like beer. I like winning. I like a campaign ecosystem where the only question anyone is asking is whether it will 'turn out voters,' not whether any of it is true."