Generative AI-produced advertising creative is now performing as effectively as human-created ads in real-world digital campaigns, according to the latest report from Rediffusion Red Lab. The study, titled “GenAI Ads Perform As Well,” analysed performance data from live advertising environments and found that AI ads delivered similar engagement and click-through rates compared to human creative across millions of ad interactions.
The research team included academics and practitioners from Columbia Business School in collaboration with researchers from TUM School of Management, Harvard Business School and Carnegie Mellon University, using real campaign data rather than controlled lab tests to gauge effectiveness.
Study scope: live campaigns and millions of impressions
The report examined more than 300,000 advertisements, spanning sectors such as personal finance, food and beverage, and education. These ads generated over 500 million impressions and 3 million real user clicks in live campaign settings, where AI creative and human creative were served side-by-side to matched audiences with identical campaign objectives and landing pages — the only variable being how the ad creative was produced.
Under these real-world conditions, the analysis found that click-through rates for AI-generated and human-produced ads moved in close alignment, suggesting AI creative has reached a level of functional maturity that supports performance-driven advertising, not just experimental novelty.
Viewer perception more important than actual origin
A key finding highlights the role of audience perception: how viewers interpret an ad’s “human-ness” had more impact on engagement than whether the creative was truly made by AI or by people. Nearly half of the ads generated by AI were perceived by users as human-made, underscoring how visually indistinguishable machine-made creative has become in real feeds.
According to the report, ads that audiences believed were created by humans earned significantly higher engagement rates, with click-through rates reaching up to 1.55%, even if those ads were actually produced by AI tools. Conversely, ads that viewers perceived as AI-generated saw engagement dip to as low as 0.20% CTR, regardless of their true origin.
Human elements still drive engagement
The study further identified visual authenticity cues, especially the presence of clear human faces, as key performance drivers. Ads featuring human faces consistently attracted stronger engagement and trust signals, while those with overly stylised visuals, heavy colour saturation, or excessive symmetry were more likely to be perceived as artificial and underperformed.
Strategic takeaways for advertisers
The Rediffusion Red Lab report concludes that AI has crossed a threshold where it can serve as a commercially viable creative engine, capable of scaling ad production without compromising effectiveness. However, the authors caution that automation alone does not constitute a creative strategy.
Across the findings, the report emphasises a simple strategic message for advertisers: AI tools can accelerate creativity, but human creative judgement remains essential. AI should be used to support emotionally resonant, authentic ad experiences rather than replacing thoughtful storytelling or flooding digital channels with technically flawless but emotionally hollow content.
As brands increasingly adopt generative technologies, the report suggests the future of advertising will likely blend AI-enabled efficiency with human-driven creativity and perceptual understanding, ensuring that ads resonate deeply with real audiences.
