AI Ads Fool Most: Can You Spot the Fake?

Detecting AI-generated advertisements is becoming increasingly difficult, as evidenced by a recent study where only 17 percent of consumers could correctly identify AI-created ads. Subtle visual inconsistencies, such as products with nonsensical parts or characters that change appearance between shots, often reveal AI's involvement. For instance, an REI ad featured a bicycle with extra handlebars, and Coca-Cola's holiday commercial showed inconsistent vehicle designs.

Further clues include unnatural movement, like wheels not rotating while a vehicle is in motion, or mismatched lip-syncing in synthetic human advertisements. Small details such as lettering on products, background elements, and even typography can also betray AI generation, though newer tools are improving in these areas. Some AI ads lean into absurdity, creating fantastical scenarios to promote products, exemplified by Fanta's Halloween campaign.

Transparency is growing, with platforms like Meta introducing labels for AI-generated or significantly altered ads. Legislation is also evolving, with Pennsylvania enacting a law requiring disclosures for AI content that creates a false impression of authenticity. As AI technology advances, identifying these ads will continue to be a moving target.