I remember the first time I realized the person on TV wasn’t real. You paused, rewound, and felt a small, uneasy click in your chest. That moment—when a face becomes a question—is where influence begins.
I tracked the numbers so you don’t have to start from scratch: political campaigns and outside groups have poured $137 million (€126 million) into ads that were created at least partly with AI tools ahead of the 2026 midterms, according to the Wesleyan Media Project. That figure is a hard line from the 2024 cycle, when researchers identified less than $50,000 (€46,000) in AI-created political ad spending.
You scrolled past an ad with a familiar face and a jarring message.
The Wesleyan Media Project has flagged 220 AI-related political ads running across 34 states this year, and I want you to feel how fast this changed. Only 30 states require disclosures about synthetic content, and researchers say that just 27% of the ads they identified included a disclaimer admitting AI was used. Erika Franklin Fowler, co-director of the project, noted the jump publicly and warned that much of the AI use is undisclosed.
How can I tell if a political ad is AI-generated?
If you want to be sharper than the feed, start with three checks I use: watch for awkward lip-sync or lighting, search for the clip on official campaign channels, and scan the post for a disclosure or a statement from a state elections office. Platforms such as Facebook, YouTube, and TikTok show different labeling behavior, and sometimes ads are thin edits of real footage created with generative tools from OpenAI, Midjourney, or video services like Synthesia to cut production costs. When in doubt, a quick reverse-image or reverse-video search often exposes recycled clips or stock footage.
A local candidate appears beside a national figure in a split-second stitch.
Nearly half of the identified AI-generated spots feature deepfake depictions of real politicians. Campaigns are constructing manufactured connections—local contenders are visually linked to national Democrats such as Hillary Clinton, Nancy Pelosi, and Alexandria Ocasio-Cortez—and those manufactured ties are running as persuasive shorthand. These ads are a fog settling over the truth, masking provenance and intent for viewers who only glance for a second.
Are AI-generated political ads legal?
Short answer: it depends on where you stand. Federal rules are still catching up and state laws vary widely. The New York State Democratic Party filed a formal complaint against Bruce Blakeman’s gubernatorial campaign, arguing his social posts violated a state law requiring disclosures for deceptive AI-generated political content; Blakeman’s team has defended the material as satire. With only a patchwork of disclosure mandates and varied platform enforcement, legality often becomes a matter of local election law and whether a regulator wants to pursue enforcement.
You notice the partisan tilt in spending; the numbers make it obvious.
Republicans are driving most of the visible spend: 81% of the identified AI ads were sponsored by Republican campaigns or groups, accounting for 88% of the dollars reported. That includes at least 22 AI-generated ad posts tied to Bruce Blakeman’s campaign in New York. Democrats, meanwhile, account for roughly 12% of the ads—part of a cautious posture after party leaders and advisers warned about reputational risk for candidates who advocate regulation while using the same technology. Travis N. Ridout, co-director of the Wesleyan Media Project, framed that tension when speaking to The New York Times.
Some Democrats are weaponizing their caution into campaign messaging. Texas Democratic Senate candidate James Talarico ran an ad that ends with the claim “Made by humans. Not AI.” That line is a rarity among political creative this season.
There is a practical side to this surge, too. Campaigns are using generative models for cheap stock footage, voice cloning, and rapid ad iterations—tools that shrink timelines and costs for small teams. When campaigns send polished, fabricated moments into feeds as Trojan horses, a short, well-targeted spot can reshape perception before a correction gets traction.
The clock is short and more AI spots are inbound.
With the midterms just over two weeks away, expect a final wave of synthetic ads aimed at voters’ timing and emotion. The Wesleyan team’s count is almost certainly an underestimate; detecting AI in creative is often a forensic effort. If you want to be strategic about what you consume, check state disclosure portals, follow reputable trackers such as the Wesleyan Media Project, and pay attention to who is paying to promote a message—not just who’s on screen.
I’m asking you to pay attention not out of alarmism but so you can vote with clearer sight. Will we let campaigns rewrite the visual rules of persuasion without stronger guardrails?