AI Deepfake Faces Rated More Real Than Photos

A University of Florida study found adults of all ages rated AI-generated faces as more realistic than actual photographs while brain activity was tracked.

Updated on Nov. 17, 2026 in Artificial Intelligence

AI Deepfake Faces Rated More Real Than Photos

Live Poll

Do you trust that you can tell real online faces from AI-generated ones?

University of Florida researchers found that AI-generated deepfake faces are rated as more realistic than real photographs. The study assembled 80 face images, half real and half AI-generated, and had 126 participants aged 18 to 89 rate their reality while undergoing brain scans.

Why it matters

The findings suggest people may be especially vulnerable to deepfakes because the faces appear more convincing than real ones, which could matter for protecting against deepfake-enabled scams. Awareness of this hyperrealism bias may be an immediate safeguard for Gainesville residents and others encountering synthetic imagery.

The study tested 80 face images, half real and half AI-generated, with 126 adults aged 18 to 89 rating their realism during brain scans. Whether the lab-based ratings reflect how people judge deepfakes in everyday online settings has not been established.

The players

University of Florida

A public research university in Gainesville, Florida, whose researchers conducted the study on deepfake face perception.

The details

Participants rated the reality of faces while anterior insula activity was tracked with brain scans, and activity in that region was lower when they viewed deepfakes. Researchers attribute the effect partly to AI faces being created as an average of many different faces, which are easier to process and more pleasing to the brain.

Timeline

  1. - The findings were published in Social Cognitive and Affective Neuroscience on November 16, 2026.

Tech Race

The study places deepfake realism within the ongoing effort to detect synthetic media, as University of Florida researchers link hyperrealistic AI faces to measurable brain responses. Its findings suggest detection challenges may grow as generative tools make faces look more convincing than real photographs.

Recognizing that AI-generated faces can appear more realistic than real photos may help people pause before trusting unfamiliar profile pictures or videos in scam-prone situations. The research gives everyday users a concrete reason to scrutinize synthetic faces in online messages and calls.

The takeaway

Being aware that AI-generated faces can seem more realistic than real ones is a practical first defense against deepfake scams. Pausing to question unfamiliar faces online is a simple habit that can help protect you.

Further reading

Explore more Artificial Intelligence coverage for Gainesville readers.

Source note: This article includes information reported by Eurasia Review.

Live Poll

Do you trust that you can tell real online faces from AI-generated ones?