In an online quiz published by Liftup, an online marketing company, participants were shown 25 images in sequence and had to decide whether each was a real photograph or generated by artificial intelligence. Viewers were given 10 seconds to evaluate each image.
Results
Test-takers were more likely to correctly identify real photos: respondents accurately labeled photographs in roughly two-thirds of cases. By contrast, about half of the AI-generated images were judged to be genuine by participants.
Why this matters
The distribution of answers indicates considerable uncertainty in human visual perception when distinguishing between synthetic and real imagery. This finding is relevant for advertising, social media, and any context where assessing the authenticity of visual content is important.
Summary
Liftup’s quiz demonstrates that many people find it difficult to reliably tell AI-generated images apart from real photographs: while real photos were correctly identified about two-thirds of the time, AI images were perceived as real in roughly half of instances.
