India’s CERT-In Urges Rapid Patching of Exposed System Vulnerabilities

CERT-In published a blueprint warning that generative and agentic AI are accelerating attack surface mapping, exploit generation, and large-scale social engineering, and mandates rapid, risk-prioritized patching (ranging from immediate/12 hours to 5 days depending on severity). The advisory also directs organizations to modernize SOC capabilities (behavioral detection, threat hunting), adopt SBOM/AIBOM for supply-chain transparency, enforce MFA and Zero Trust, and defend AI systems against adversarial techniques like prompt injection and training-data poisoning.