Cert-In issues AI security audit guidelines, sets 4-tier risk framework

CERT-In seeks checks on access of AI agents

NEW DELHI: India's cybersecurity agency has called for testing whether AI agents can leak confidential data or take unauthorised actions, putting

CERT-In calls for testing AI agents for unauthorised tool use, privilege abuse

Guidelines require checks on agents’ permissions, task delegation and emergency-stop mechanisms; multi-agent systems must maintain authenticated, logged communications and records of actions

Cert-In issues AI security audit guidelines, sets 4-tier risk framework

Cert-In emphasised that compliance requirements for high-risk and critical systems apply uniformly, with no exemptions or relaxed standards for smaller firms or startups.

CERT-In’s AI Security Auditing Guidelines set out tiered framework

The 9 core audit domains outlined 'AI Security Audit Guidelines' not just covers AI systems developed, procured and deployed or operated in India, it also includes those processing Indian data or providing AI-enabled services to Indian users.