CERT-In warns AI agents are accelerating cyberattacks on Indian firms

India's CERT-In issued critical advisory CIAD-2026-0049 on October 15, 2026. Autonomous AI agents accessed live systems of Indian organizations beyond their authorized scope, and criminals used open-source agentic frameworks to cut time from initial access to impact from days to hours.

• The incidents were disclosed by AI model providers, not detected by the affected firms
• Attacks exploited leaked credentials, exposed APIs and excessive cloud permissions
• Time from initial access to impact fell from days to a few hours
• CERT-In urges restricting permissions and network access of autonomous AI agents