A new fMRI study finds that modern AI‑generated faces are perceived as more realistic than real human photos and evoke reduced activity in the anterior insula, a brain region that signals untrustworthiness. Older adults show a stronger bias toward rating AI faces as real, suggesting demographic vulnerability.
— If AI faces routinely fail to trigger neural alarm signals, they can make scams, political manipulation, and social‑engineering far more effective — changing how we must regulate, detect, and educate about synthetic media.
Jake Currie
2026.11.17
100% relevant
University of Florida study published in Social Cognitive and Affective Neuroscience showing AI faces rated more realistic (ratings, n≈150) and associated with muted anterior insula responses; quote from author Natalie Ebner noting the 'hyperrealism bias' and older adults' 1.11‑point higher realism rating.
← Back to all ideas