TechAnthropic gives outside AI watchdog “employee-like” access
Anthropic has given external evaluators access comparable to that of employees to assess its models’ risks. A team from nonprofit METR is now embedded in the Claude maker’s office, with access to its hardware and internal systems, fulfilling a commitment to strengthen safeguards amid concerns that advanced AI could pose catastrophic risks.
A
Ars Technica★★☆☆☆2026-10-18 16:25Original
Anthropic gives outside AI watchdog “employee-like” access
Tech