# Would YOU trust these people? Fake faces created by AI look MORE trustworthy than real people, study reveals - take the test to see if you can tell the difference > Source: > Published: 2026-07-10 14:08:17+00:00 # Would YOU trust these people? Fake faces created by AI look MORE trustworthy than real people, study reveals - take the test to see if you can tell the difference **READ MORE:**[Scientists reveal method for spotting AI-generated faces](/sciencetech/article-15937659/distinguish-AI-generated-faces-test.html)**See more Daily Mail on Google -**[save us as a Preferred Source](https://google.com/preferences/source?q=dailymail.com) Can you tell the difference between a real face and one generated by [AI](/sciencetech/ai/index.html)? According to a new study, it might be a lot harder than you think. Researchers at Lancaster University discovered that people are no better than chance at distinguishing AI imposters from the real deal. To make matters worse, people actually find AI-generated faces more trustworthy than their human counterparts. According to the researchers, this creates a serious risk that people could fall victim to AI-powered identity fraud or catfishing. Lead author Alexis McGuire, a PhD student at Lancaster University, told the Daily Mail: 'The fact that people often [perceive AI-generated faces as trustworthy](/sciencetech/article-10515187/Fake-faces-created-AI-look-trustworthy-real-people-study-reveals.html) makes them particularly powerful tools for online scams and disinformation. 'For example, a text-based scam may become more convincing if it is accompanied by a face that people instinctively trust.' So, how many of these AI-generated faces can you distinguish from real people? Take the quiz below to find out. AI-generated faces used to be fairly easy to pick out due to 'AI artefacts' like sixth fingers, misaligned teeth, or wonky ears. However, new research suggests that this advice[ doesn't improve people's ability to spot deepfakes](/sciencetech/article-15283159/spot-AI-fake-face-test.html), and real-life fraudsters can easily edit out or avoid these errors. Likewise, this new study suggests that the latest generation of image-generation models have[ become almost impossible for humans to detect](/sciencetech/article-15937659/distinguish-AI-generated-faces-test.html). 'If people don’t continually update their knowledge about what to look for then it can give a false sense of security and make them more, not less, vulnerable,' says Ms McGuire. In their paper, published in the [Journal of Vision](https://jov.arvojournals.org/article.aspx?articleid=2818019), scientists tasked 169 participants with assessing a collection of 96 real and fake faces. Participants were shown a randomly selected face and asked to select whether it was AI-generated or real. However, on average, people were only correct 58.4 per cent of the time - barely better odds than a coin flip. The accuracy varied by ethnicity and the specific AI model being used, but the trend was largely consistent. Scientists found that most people are only just better than chance when it comes to spotting the difference between real (middle) and AI-generated (left and right) faces Surprisingly, the researchers found that the faces made by the newest 'diffusion model' AIs were easier to spot than those made by the older 'generative adversarial network' (GAN) models. But the most shocking result of all came from a follow-up test, in which participants were asked to rate the trustworthiness. Real human faces were consistently rated as the least trustworthy, scoring just 4.04 on a scale from one to seven - with seven being very trustworthy. The exceptionally hard-to-spot GAN faces were rated more trustworthy, with a score of 4.36, while the diffusion model faces were the most trusted of all at 4.7. Strangely, this means that people trusted faces created by a diffusion model more even though they thought they were less realistic. Ms McGuire says: 'This finding presents a paradox and thus highlights the possibility that realism and trustworthiness judgements are driven by two different psychological mechanisms.' The researchers suggest that this could be because AI-generated faces tend to cluster towards the 'average' human face. When we encounter certain types of facial features frequently, our brains cluster these details together to form a representation of what a face should look like. Which of these faces do you find more trustworthy? Scientists found that AI-generated faces (left) were rated as more trusted than real faces (right) New faces are assessed in relation to that cluster, and the closer they fall to the average, the more familiar they feel. Since AIs aggregate millions of humans into an average mixture, this process might make them seem more typical and trustworthy than a real person; but this is unlikely to be the whole story. AIs tend to create 'polished, idealised faces that appear especially attractive', which people tend to instinctively find appealing. Ms McGuire adds: 'They have features that people naturally associate with trust, such as being more attractive. 'Research has long shown that people often perceive attractive individuals as more trustworthy.' That creates a serious worry that AI-generated faces could become a perfect tool for fraudsters and criminals looking to gain their victims' trust. If you want to take part in the research, the University of Lancaster has created an online survey, available [at this link](https://run.pavlovia.org/lexmcguire/study_v13_clean/), where you can test your ability to spot AI faces.