RealFeedApp External auditors received access to Anthropic hardware
Nonprofit METR has been granted access to Anthropic's hardware and internal systems to assess AI model risks. This marks the first step in implementing proposals for embedding external auditors in leading labs.
- Field
- how neural networks work
- What they did
- Anthropic granted external evaluators access to its internal systems and hardware to assess the risks posed by its AI models.
- Why it matters
- This step aims to verify the safety of advanced AI systems, as uncontrolled behavior could threaten global economy, cybersecurity, and humanity.
#anthropic#metr#ai safety#external auditors
