Anthropic says it stopped scientists potentially developing bioweapons with AI
Start-up discloses five examples of when users ‘circumvented controls’ and made efforts to ‘obfuscate’ the purpose of their research
TL;DR
- Anthropic, an AI start-up, identified five cases of user circumvention of controls.
- Users made efforts to obscure the purpose of their research.
- The disclosed research involved potential bioweapon development.
- Anthropic successfully stopped these attempts.
- The disclosure highlights the risks of AI misuse.