Disclosing AI has become part of the brand: what election ads teach us

The US midterm campaign has become the largest public test of advertising made with artificial intelligence. And the most interesting result is not the volume. It is the silence.

On October 15, 2026, the Wesleyan Media Project, a Wesleyan University center that tracks political ads, published its latest count: 220 ads made fully or partly with AI, with more than $137 million in airtime. Only 27% of them told viewers there was AI in the ad.

In other words: out of every 10 films, 7 kept quiet.

How the number grew

Wesleyan has been measuring this throughout the year. In May, there were 33 ads and about $20 million. In the September 4 report, there were already at least 164 ads and almost $80 million, based on AdImpact estimates. At that point, 69% of the ads and 78% of the money spent carried no disclaimer at all.

The researchers themselves treat the numbers as a floor, not a ceiling. The reason is simple: an ad that discloses AI is easy to find. One that does not may go unnoticed.

There is no single rule either. In the United States, no federal law requires a disclaimer on AI-made political ads, as Wesleyan notes in its September report.

The other side: reputation

For anyone who works with brands, the most revealing figure is not the money. It is the choice not to disclose. Whoever produces an ad knows whether they used AI. If 7 out of 10 chose not to say so, it is because they bet the audience would react badly to the information.

In other words, the market already understands that AI affects trust. Some deal with it by hiding. Others by avoiding the tool. Almost no one is dealing with it the simplest way: owning, clearly, what was done and why.

One caveat: a label, on its own, does not solve the problem. The core problem is deception, not the technology. A disclaimer does not save an ad made to deceive. And an honest ad is not made worse by having one.

In Brazil and Europe, disclosure is already the rule

In Brazil, the debate arrived earlier. On March 2, 2026, the Superior Electoral Court (TSE) approved the rules for the general election, whose first round was on October 4. The campaign advertising rule forbids spreading synthetic content, generated or modified by AI, outside the labeling rules.

In the European Union, Article 50 of the AI Act deals with transparency. Anyone who uses AI to generate or alter images, audio or video that imitate real people, places or events (the so-called deepfake) must disclose that the content was generated or manipulated. Under the regulation’s timeline, these obligations have applied since August 2, 2026.

An election is not a brand campaign, and the rules are different. But the direction is the same in all three places: audiences want to know when they are looking at a fabricated image. Brands that advertise in Brazil, in Europe or both will meet that expectation in more and more places.

What changes for brands and agencies

The disclaimer is usually treated as a footnote, added at the end, out of obligation. This post argues the opposite: disclosing AI is a brand decision, and it should be made in the brief, not at delivery.

In practice, that means:

• Decide early. The disclaimer goes into the script and the budget, like music and subtitles. Decided at the end, it becomes a makeshift badge.
• Be specific. “Made with AI” says little. “Sets and voice generated with AI; direction, script and editing by people” tells the audience what it needs to know.
• Put it where it can be seen. In the ad, in the post caption and on the product page. A hidden disclaimer works like no disclaimer.
• Separate craft from deception. An impossible set, animation, an invented world: that is film language. A real person saying what they never said, a simulated testimonial, a product that does not exist: that is deception, with or without a label.
• Keep the record. Which tools, which versions, who approved. If anyone asks, the answer is ready.

None of this replaces checking the rules of each country and each platform. It is the brand layer that comes first.

Why disclosure can work in your favor

Campaigns fear the label will diminish the ad. That makes sense when the film depends on the trick to impress. When the piece stands on its idea, rhythm and character, the disclaimer becomes something else: it shows the brand knows what it did and is not afraid to say so.

It is the same point as the post on Grok’s AI-made Odyssey: the technique is not the film. What audiences remember is the story. And what they demand afterwards is honesty about how it was made.

This blog’s point of view

Wesleyan’s number does not only show a regulatory gap. It shows an opportunity. If 7 out of 10 ads hide the AI, the brand that discloses clearly stands out without spending anything extra.

That is why, in this portfolio’s AI brand films, AI use is always disclosed. Not out of obligation, but because the method is part of the conversation with the audience, just like the direction that decides every shot.