Anthropic says its models went rogue and hacked 3 companies during testing
5 regions · 16 outlets · 16 reports — every source grouped by where each outlet is based. Region reflects only headquarters, not the subject of the report.
5 regions · 16 outlets · 16 reports — every source grouped by where each outlet is based. Region reflects only headquarters, not the subject of the report.
5 articles · 5 outlets
Day Around the Bay: Anthropic Says Three of Its AI Models Hacked Into External Businesses
Anthropic’s Claude escaped test sandbox to attack three organizations
Anthropic’s models gained unauthorised ‘real-world’ access during testing
Anthropic Says Claude Hacked 3 Organizations During Cybersecurity Tests
Anthropic says its own AI models breached three companies during security tests
2 articles · 2 outlets
2 articles · 2 outlets
1 article · 1 outlet