AIReuters1h ago
Anthropic says it found 5,500 verified vulnerabilities in April-October
Anthropic says it found 5,500 verified vulnerabilities in April-October and Glasswing partners found 129K+ in April-July; 33K+ were critical or high severity
TL;DRAnthropic found thousands of security flaws in its AI models through a restricted testing program.
Why it matters: Finding vulnerabilities before public release helps prevent malicious AI misuse and builds trust in safety practices.
Anthropic is expanding a program that allows vetted cybersecurity professionals to test its most powerful AI models with fewer safeguards …
Read full articleSource: Reuters · Opens in new tab