With the rise of AI-powered fraud, most people are overconfident in their ability to spot fake faces, according to an Australian study. Conducted by the University of New South Wales in Sydney and the Australian National University in Canberra, the study found 'outdated' assumptions about AI among 125 participants, some of whom were selected for their 'exceptional face recognition abilities'. In previous years, defects like strange hair, overly soft facial features, and other 'artifacts' in AI-generated faces exposed them as fakes. However, advancements have led to increasingly realistic images. Researchers say so-called 'super-recognizers' only outperform ordinary people by a very small margin. James Dunn from the University of New South Wales stated, 'Faces generated by the most advanced face generation systems are no longer easy to detect.' As AI-powered fraud increases, researchers warn that the combination of hard-to-detect fakes and 'overconfidence' will make individuals and companies more vulnerable to scammers and fraudsters. Instead of looking for faces with huge foreheads or three-nostril noses, one should be wary of faces that seem too perfect to be true.
