AITechCrunch AI2h ago
Anthropic can’t reliably control its AI agents. It’s cutting off its
Anthropic can’t reliably control its AI agents. It’s cutting off its internal evals from the live internet instead
TL;DRAnthropic disabled live internet access for its AI testing due to control concerns.
Why it matters: Reveals safety challenges with AI agents and raises questions about deployment readiness.
Anthropic said it "turned off live internet access" for "all our internal evaluations" until further notice.
Read full articleSource: TechCrunch AI · Opens in new tab