05-08-2026 11:24
via
bbc.co.uk
Anthropic AI created fake profiles and impersonated people in attempted hack
The UK's AI Safety Institute said recent behaviour from Anthropic and OpenAI models was malicious and unprecedented.
Read more »