UNSW Face Test

Why this matters?

AI-generated faces are becoming increasingly realistic and widely accessible. These synthetic images are now used in social media profiles, marketing materials, political messaging, and even fraudulent schemes. While some uses are harmless, others can contribute to identity deception, misinformation, financial scams, and the manipulation of public opinion. As the technology improves, it becomes harder to distinguish real faces from artificial ones. We created this test to show how hard this task has become. The average person does no better than guessing when distinguish real from AI, but some people are super-AI-detectors.

By studying individual differences in detecting AI-generated faces, we hope to understand who is most vulnerable to deception and how to design tools and educational strategies that strengthen digital resilience.