Skip to content
Science & Technologyv1 · 1 version

CERT-In's AI Security Audit Guidelines

The AI Security Auditing Guidelines are CERT-In's framework for checking whether an AI system is secure, fair, explainable and properly governed, from design to retirement. Version 1.0, dated 4 November 2026, grades AI systems into four risk tiers and sets audit rules for agentic and other advanced AI.

Last updated 8 November 20264 min read
Latest update
New card: CERT-In's AI audit framework of 4 Nov 2026, with its four risk tiers, nine audit domains, agentic-AI controls and who must be audited by empanelled auditors.

Read the full living note — free

The complete note in both exam lenses (UPSC analytical / fact capsule), its full version history, and “what changed since you last read” — plus tonight's quiz.

Open in CA·Wiki

AI-assisted, human-reviewed · sourced from PIB, RBI & ministries · editorial policy

More living notes in Science & Technology

Browse all topics →

CERT-In's AI Security Audit Guidelines — Science & Technology current affairs | CA·Wiki