Google’s generative artificial intelligence bot Gemini is potentially being used to create deepfake child pornography as well as terrorist and violent extremist material, according to data in a report from the eSafety Commissioner.

The report reveals that Google has received 86 reports from users around the world about suspected synthetic child sexual exploitation and abuse material generated by the AI text and image generator between April 2023 to February 2024. Over the same period, Gemini users reported 258 instances of potential deepfake terrorist and violent extremist material.