CERT-In gets access to Anthropic’s Mythos AI to test cyber vulnerabilities

Agency to examine AI model’s ability to detect and potentially exploit security flaws in controlled environment following incidents involving unauthorised access to Indian companies’ systems