Skip to main content

CSCRF + AI Advisory · May 2026 · Combined Reference

SEBI CSCRF + AI Advisory Combined Readiness Guide

On 5 May 2026, SEBI issued the AI Vulnerability Detection Advisory (circular HO/13/19/12(1)2026-ITD-1_CIMGI/10873/2026). It must be read in conjunction with CSCRF. This guide maps all 10 Annexure-A directives to CSCRF controls, lays out per-tier readiness, and gives you a 90-day sequence.

The 10 directives → CSCRF controls

Each Annexure-A directive maps to one or more CSCRF control families. The "tier impact" column indicates which tiers carry binding obligations, and the "evidence" column tells you what your auditor will ask for.

1

Patch immediately + virtual patching

Update all OS and applications with latest patches immediately. Where patches are unavailable, adopt virtual patching (WAF, ModSecurity, network ACLs) as an interim measure.

CSCRF Control
PR.MA (Maintenance) + DE.CM (Continuous Monitoring)
Tier Impact
All tiers. MIIs + QREs: half-yearly VAPT surfaces patch gaps; virtual patching bridges the window between discovery and vendor release.
Audit Evidence
Patch matrix with date-applied column. Virtual patching rationale for unpatchable findings.
2

Vulnerability assessment with AI-based VA tools

Conduct VA using conventional AND suitable AI-based tools where possible on a regular/continuous basis per CSCRF cadences.

CSCRF Control
DE.CM + ID.RA (Risk Assessment) + VAPT framework
Tier Impact
All tiers. MIIs + QREs: AI-based VA tools should be integrated into the twice-annual/annual VAPT cycle. Mid/Small/Self-cert: evaluate adoption; document rationale.
Audit Evidence
AI VA tool evaluation memo. VAPT report referencing AI-based tooling used.
3

Third-party + COTS vendor risk (shall)

Engage third-party vendors to release timely patches. Exchanges and Depositories shall direct empaneled application vendors to assess AI-led vulnerability detection risks.

CSCRF Control
GV.SC (Supply Chain Risk Management)
Tier Impact
MIIs + QREs: binding. Mid-size: IT Committee consultation. Exchanges/Depositories: additional obligation to direct COTS vendors.
Audit Evidence
Vendor risk assessment letters. COTS vendor AI-risk assessment responses.
4

Change management (including minor changes)

Any change — including minor changes — must encompass full documentation, impact analysis, review, testing, and secure deployment.

CSCRF Control
PR.IP (Information Protection) + ID.RA
Tier Impact
All tiers. Broader than CSCRF material-change framing. Every change — including config changes, hotfixes, dependency bumps — needs documented change control.
Audit Evidence
Change management register. Impact analysis and rollback plan per change.
5

API security

Maintain API inventory. Enforce strong authentication/authorisation with least privilege. Implement rate limiting and throttling. Whitelist-based API connections.

CSCRF Control
PR.AA (Access Control) + PR.DS (Data Security)
Tier Impact
All tiers. MIIs + QREs: full API inventory and threat modelling. Mid/Small/Self-cert: inventory + authN/Z review.
Audit Evidence
API inventory document. Rate-limit config. AuthN/Z audit trail.
6

SOC monitoring + SOAR/SIEM + M-SOC onboarding (shall)

All eligible REs shall expedite M-SOC onboarding. Examine ALL SOC alerts including low-priority. Integrate SOAR playbooks with SIEM. MIIs: run awareness workshops for member onboarding.

CSCRF Control
DE.CM + DE.DP (Detection Processes) + M-SOC (CSCRF §4.5)
Tier Impact
Small-size + Self-cert: M-SOC mandatory — this is the most time-sensitive obligation. QRE + Mid-size: eligible — accelerate onboarding. MIIs: run workshops.
Audit Evidence
M-SOC onboarding status. SOAR playbook library. SIEM alert-triage log.
7

CSCRF risk assessment with AI-based threat scenarios

Include AI-based model capability as a risk scenario in every CSCRF periodic risk assessment. Speed-and-scale exploitation, agentic chain abuse, prompt injection.

CSCRF Control
ID.RA (Risk Assessment)
Tier Impact
All tiers. MIIs: half-yearly risk assessment includes AI scenario. QRE + Mid-size: annual assessment includes AI scenario. Add AI-attacker scenario to risk register.
Audit Evidence
Updated risk register with AI-threat entries. Risk assessment report referencing AI scenarios.
8

Zero Trust + system hardening

Implement ZTNA architecture. Apply CIS-style hardening — secure configurations, disable unnecessary services and default accounts, least privilege. Network segmentation, no SPOF, HA.

CSCRF Control
PR.AA + PR.AC (Access Control) + PR.PT (Protective Technology)
Tier Impact
All tiers. Aug 2025 reframed ZT as methodology-driven ("approved by IT Committee") rather than prescriptive.
Audit Evidence
ZTNA architecture documentation. System hardening baseline. IT Committee approval record.
9

SBOM + asset inventory

Periodically update asset inventory and Software Bill of Materials for all critical applications — including open source stack.

CSCRF Control
ID.AM (Asset Management)
Tier Impact
All tiers. First SBOM mandate in Indian sectoral regulation. Critical applications only (as defined per CIR/2025/119 — same network segment).
Audit Evidence
Asset inventory. SBOM per critical application. Open-source dependency scan report.
10

IT-committee guidance + long-term AI plan (shall)

MIIs and REs shall seek IT-committee guidance. All REs must prepare a long-term plan for AI in detection and autonomous/agentic mitigation. Recalibrate risks. Plan AI-augmented SOC transformation and continuous VA using AI tools.

CSCRF Control
GV.PO (Policy) + GV.OV (Oversight)
Tier Impact
Binding for ALL tiers. MIIs + QREs + Mid-size: IT Committee must ratify. Small + Self-cert: MD/CEO/Board must approve. Deliverable: a documented long-term AI strategy.
Audit Evidence
IT Committee minutes addressing AI plan. Long-term AI strategy document. Risk recalibration memo.

Per-tier readiness roadmaps

MII

Priority

Highest. Items 3, 6c, 10: binding shall language. M-SOC workshops (item 6c) are your additional obligation. COTS vendor direction (item 3).

Timeline

Weeks 1-4: gap assessment + AI VA tool evaluation. Weeks 5-8: SOAR/SIEM integration + M-SOC workshop design. Weeks 9-12: long-term AI plan + IT Committee submission.

Qualified RE

Priority

High. Items 3, 6c, 10: binding. Item 7: AI scenario in risk assessment. Item 2: evaluate AI VA tools. M-SOC onboarding acceleration.

Timeline

Weeks 1-4: AI VA tool evaluation + vendor risk letters. Weeks 5-8: M-SOC onboarding + SBOM generation. Weeks 9-12: AI plan + risk recalibration.

Mid-size

Priority

Moderate. Items 6, 7, 10: primary. IT Committee must ratify AI plan (item 10). Item 2: evaluate AI VA tools. Item 9: SBOM for critical apps.

Timeline

Weeks 1-4: asset inventory + SBOM + AI VA evaluation. Weeks 5-8: AI risk scenario in assessment + SOAR integration. Weeks 9-12: IT Committee AI plan submission.

Small-size + Self-cert

Priority

Focused. Item 6c: M-SOC onboarding is TIME-SENSITIVE. Item 10: AI plan (MD/CEO approval, not IT Committee). Item 7: AI risk scenario.

Timeline

Weeks 1-4: M-SOC onboarding kickoff + asset inventory. Weeks 5-8: SBOM + API inventory + AI risk scenario. Weeks 9-12: AI plan submission.

90-day readiness checklist

Weeks 1–4: Gap assessment

☐ Map current cyber posture against 10 Annexure-A directives

☐ Inventory APIs, COTS vendors, AI/ML pipelines, and SOC log sources

☐ Identify M-SOC onboarding eligibility and gaps

☐ Surface long-term AI plan stakeholders (IT Committee, CISO, board)

☐ Generate SBOM for critical applications

☐ Confirm CSCRF tier classification (use wizard → /compliance/sebi/wizard)

Weeks 5–8: Implementation

☐ AI-augmented VAPT cycle (B-52 or equivalent AI-based VA tool)

☐ Third-party vendor risk assessment letters

☐ API inventory completed + authN/Z audit

☐ M-SOC onboarding: log-source inventory + SIEM integration

☐ SOAR playbooks designed and tested

☐ AI-attacker scenario added to risk register

Weeks 9–12: Submission

☐ Long-term AI plan drafted (Annexure-A item 10 deliverable)

☐ AI plan ratified by IT Committee (or MD/CEO for Small/Self-cert)

☐ Risk register recalibrated for AI-accelerated threats

☐ SOAR/SIEM integration validated

☐ Evidence pack compiled for every Annexure-A item

☐ Board / IT Committee submission packaged

Ready to scope your combined CSCRF + AI Advisory readiness?

Security Brigade has been CERT-In empanelled since 2008. We deliver B-52 powered VAPT, AI-system-defender testing, M-SOC readiness support, and Annexure-A evidence packs.

Source. CSCRF: SEBI/HO/ITD-1/ITD_CSC_EXT/P/CIR/2024/113 (20-Aug-2024) read with amendments. AI Advisory: HO/13/19/12(1)2026-ITD-1_CIMGI/10873/2026 (5-May-2026). Reviewed: 2026-05-06. Informational, not legal advice.