Artificial intelligence is changing how political campaigns compete for voters in the United States, and not every campaign is playing by the old rules of political advertising.
As the country prepares for its November 3, 2026, midterm elections, candidates and their supporters are increasingly using AI to create videos showing political opponents in scenes that never happened. Some are designed as obvious jokes, while others risk leaving voters confused about what is real.
The technology makes it possible to produce attention-grabbing political ads faster and at lower cost than traditional production methods. But as AI-generated content spreads, a bigger question is emerging: can voters trust what they see during an election campaign?
AI Political Ads Are Becoming a Major Campaign Tool
The scale of AI use in this election cycle is already significant. The Wesleyan Media Project, which tracks political advertising, identified at least 164 ads created or enhanced using AI, with estimated spending just under $80 million in its September report.
Its findings showed that Republican candidates and pro-Republican groups accounted for about 80 percent of the identified ads. The researchers cautioned that the figures were an undercount because some uses of AI are difficult to detect.
One example involves Bruce Blakeman, a Republican candidate for governor of New York. His campaign released AI-generated ads depicting Democratic figures Zohran Mamdani and Kathy Hochul in fictional situations, including a scene suggesting they were working together in ways that would increase costs for residents.
Neither politician actually appeared in the generated scenes.
Other ads have used AI to place candidates in unusual or exaggerated situations. Some portray political figures as fictional characters or use comedy to criticise their records.
Supporters of this approach argue that obvious satire is part of American political tradition. Critics, however, warn that even an exaggerated or clearly artificial scene can influence how voters perceive a candidate.
The Rise of ‘Slopaganda’
A term gaining attention in the debate is “slopaganda,” which combines low-quality AI-generated content with political propaganda.
Unlike a convincing deepfake designed to make viewers believe a fabricated event really happened, slopaganda can be deliberately absurd. Its purpose may be less about proving a specific claim and more about shaping a candidate’s public image.
For example, a fictional video showing a politician behaving selfishly might not be intended to fool viewers into believing the scene literally happened. Instead, it could reinforce the message that the politician is out of touch or untrustworthy.
That distinction matters because political advertising can influence opinions through humour, repetition and emotional reactions—not just through factual claims.
AI also allows campaigns to experiment with different versions of an ad without the expense of organising a real-world production. This makes the format attractive to political strategists competing for attention across television and digital platforms.
However, the impact on voter decisions is still difficult to measure. A video can attract millions of views without necessarily changing how people vote.
Why AI Labels and Election Laws Matter
One major concern is that voters may not always know when an advertisement has been created or altered using AI.
The United States does not have a comprehensive federal requirement covering AI disclosures across all political advertisements. Instead, states have introduced different rules, creating a patchwork of requirements that can vary by location and type of election.
The Wesleyan Media Project found that 69 percent of the AI-related ads it identified did not carry a disclaimer disclosing AI use. The figure highlights a transparency problem, although the absence of a label does not automatically mean every ad is deceptive or illegal.
Supporters of clearer disclosure rules argue that voters should be able to identify AI-generated material before forming an opinion about a candidate. Some experts have compared the idea to food labels, which help consumers understand what they are buying.
The challenge is finding a balance between transparency and freedom of political expression. AI can be used for satire, harmless visual effects and legitimate campaign communication, so rules must account for differences between ordinary editing and fabricated content that could mislead voters.
Clear, visible disclosures could help people make better-informed decisions without banning the use of AI altogether.
A Bigger Test for Voter Trust
The controversy is also affecting candidates who use AI tools in their own campaigns.
In California, Democratic state Senator Scott Wiener faced criticism after launching an AI chatbot focused on his congressional opponent, Connie Chan. The chatbot answered thousands of questions about Chan and used parody to criticise aspects of her political record and background.
The move triggered backlash, including criticism that some of the content was offensive. Wiener eventually withdrew the chatbot, saying the effort had missed the mark.
The episode shows that using AI in politics can create reputational risks even when a campaign intends to be innovative or humorous. Voters may judge not only the content itself but also the decisions behind its creation.
As the November elections approach, AI-generated advertising is likely to remain part of the campaign landscape. The central issue is not simply whether artificial intelligence should be used in politics, but whether campaigns use it transparently and responsibly.
For voters, a practical lesson is to pause before sharing a dramatic political video, look for its original source, check whether reputable outlets have verified the claim and distinguish a clearly labelled parody from evidence of a real event.
AI can make political advertising cheaper, faster and more creative. But when fictional scenes blur the line between satire and reality, public trust becomes the real prize at stake.
