UNSW Launches AI Faces Test Amid Growing Concerns Over Deepfakes

Published on May 20, 2026

In a world filled with selfies, portraits, and online personas, distinguishing between real and artificial images has become increasingly complex. Traditional methods of identifying photography relied on human intuition and experience. However, technological advancements are now blurring these once-clear lines.

The University of New South Wales recently introduced an interactive AI faces test aimed at assessing users’ ability to discern fake portraits from genuine ones. As prominent figures like Guardian Australia’s Carly Earl and Matilda Boseley participated, the results raised eyebrows. Users quickly realized that many AI-generated faces were nearly indistinguishable from real photographs.

The test revealed that even seasoned photographers struggled to identify synthetic images. Both participants expressed surprise at the proficiency of AI in emulating realistic human features. With AI technologies becoming increasingly sophisticated, reliance on visual cues alone is proving inadequate.

This shift has significant implications for various sectors, including media and art. As the ability to create convincing fake images grows, concerns about misinformation and authenticity intensify. Users and professionals are left grappling with the challenge of trust in the digital age.

Related News