The UK’s AI Safety Institute said recent behaviour from Anthropic and OpenAI models was malicious and unprecedented.
Source: BBC.
Written by
in
