Google DeepMind study: These deepfakes are the most ‘dangerous’ form of AI use

AI-generated “deepfakes” impersonating prominent figures like politicians and celebrities are a more prevalent threat than using AI for cyberattacks, a study by Google’s AI division, Google’s DeepMind, has said.
The study, conducted alongside Jigsaw – Google’s research and development unit, analysed roughly 200 cases of AI misuse reported between January 2023 and March 2024.