We evaluate White AI faces as authentic more often than human faces.

Are you ready to be amazed? Researchers at the Australian National University recently conducted a study that revealed a startling truth – people are more likely to misidentify hyperrealistic AI faces as real than actual human faces!
The study included experiments to determine whether participants could differentiate between human and AI-generated faces and the features that led to the choices they made. The results were astonishing – AI-generated faces were misidentified as human 65.9% of the time, whereas human faces were judged as human only 51.1% of the time. Co-author Dr. Eva Krumhuber commented, “Artificial intelligence has reached an astonishing level of realism, and here we find that sometimes it can even seem more real than reality – hyperrealism – so that we can be very easily tricked into thinking an AI-generated face is real.”
Moreover, the participants who made the most errors were the most confident that they were making correct judgments. Elizabeth Miller, study co-author and PhD candidate at Australian National University said, “Concerningly, people who thought that the AI faces were real most often were paradoxically the most confident their judgments were correct.”
The researchers concluded that AI-generated faces aren’t just indistinguishable from human faces, but they have features that make them seem even more real to us. This feature of AI faces was termed ‘hyperrealism’. The AI faces were created using StyleGAN2, a generative adversarial network (GAN) trained on a large dataset of human faces with around 69% being White and 31% for all other races combined. This overrepresentation of White faces enabled the model to generate faces that represented the average of all those features in a way that made them look more human than human.
The researchers concluded that this bias in the training datasets raises important issues. “If AI faces do appear more realistic for White faces than other groups, their use will confound perceptions of race with perceptions of being ‘human’”, it noted.
Senior author Dr. Amy Dawel said, “If white AI faces are consistently perceived as more realistic, this technology could have serious implications for people of color by ultimately reinforcing racial biases online.”
Combining this data with the first experiment, the researchers were able to identify what made people more likely to identify a face as being AI-generated or real. They found that the hyperrealism of AI faces could be attributed to them being “significantly more average (less distinctive), familiar, and attractive, and less memorable than human faces.”
The fact that we’re so quick to accept that an AI face is real shows how important it is to have AI fake detection tools. The researchers used the data from the human-perceived attributes and how they were used correctly and incorrectly when misidentifying AI faces and created a machine learning model to spot AI faces. This model was able to accurately classify face type with 94% accuracy.
Dr. Dawel summed up what our best option is in the face of this: “Educating people about the perceived realism of AI faces could help make the public appropriately skeptical about the images they’re seeing online.”
So, let’s all make sure to keep our eyes open and stay aware of the potential dangers of AI-generated faces appearing online. If we remind ourselves that we’re really bad at spotting fakes, maybe we’ll be less likely to be fooled by them. Let’s take on the challenge – can you tell which of these faces are real and which are AI-generated? Test your skills and find out how you did at the end of the article!