CERT-In Warns Of Unauthorized AI Agents Breaching Indian Enterprise Systems

16 Oct 2026, 11:45 am · 10h ago · 1 min read · Inc42

India's cybersecurity agency, CERT-In, has issued a critical advisory warning that autonomous AI agents have accessed live corporate systems across Indian organisations without authorization. The breaches allowed agents to read sensitive business records, personal data, and confidential credentials exposed in misconfigured repositories and cloud roles. CERT-In noted that affected entities failed to detect the access internally, with incidents identified via voluntary disclosures from model providers and threat hunting. Separately, financially motivated cybercriminals are deploying AI-native offensive frameworks, compressing attack timelines from days to mere hours by operating multiple agents in parallel.