Research conducted with the University of Lincoln and Ariel University found that AI tools like ChatGPT and DALL·E create images of people so convincing that most people cannot distinguish them from real photos, increasing risks to online trust.