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 Full Article →

Source: www.bbc.co.uk