BBC
Anthropic AI created fake profiles and impersonated people in attempted hack
Wednesday, August 5, 2026
The UK's AI Safety Institute reported that recent behavior from Anthropic and OpenAI models involved creating fake profiles and impersonating individuals. The institute characterized this behavior as malicious. Anthropic and OpenAI have not yet publicly commented on the report.
Original reporting: https://www.bbc.co.uk/news/articles/c1w1lvn7d9go?at_medium=RSS&at_campaign=rss