ISO/IEC 42001:2023·Annex A A.6.2.8·high·Component: Technological controls

What this control requires

AI system development activities shall include controlled build, test, validation, and release practices with risk-based gates.
— ISO/IEC 42001:2023, Annex A A.6.2.8

In plain English

AI system development activities shall include controlled build, test, validation, and release practices with risk-based gates. In practice, this means you can prove you develop risk-level controls on a defined cadence.

Why it matters

ISO 42001 is the first audited evidence regulators, customers and investors can ask for that you are managing AI risk. Each Annex A control is a check-box in that audit; it is also the management system enterprises are starting to require from AI vendors before an MSA can be signed.

This control is rated high. Auditors expect a named owner, a remediation SLA, and a tooling artefact that evidences continuous operation.

What an auditor looks for

An ISO 42001 auditor verifies you have an AI policy, a risk-based system of impact assessments for each AI system, lifecycle controls from data acquisition through retirement, and third-party-AI controls that pierce the vendor's wrapper to underlying model behaviour.

Within Technological controls, the assessor will trace your implementation back to the catalogue reference (Annex A A.6.2.8) and look for objective evidence on the latest two assessment cycles.

How to implement

Secure Development

Pipeline. Require code review, SAST (Semgrep / Snyk Code), SCA (Dependabot / Renovate), secret-scanning and IaC scanning in the merge pipeline.

Branch protection. Protected branches enforce reviews, signed commits and passing CI before merge; document the protection rules as the control evidence.

Vulnerability Management

Scanners. Pair an unauthenticated scanner (Nessus / Qualys / Tenable.io) for external surface with an authenticated agent for OS/library CVEs on every host.

SLA. Document remediation SLAs by CVSS / EPSS: critical 7 days, high 30 days, medium 90 days; report monthly trend to leadership.

Containers. Scan images in CI (Trivy / Grype / Snyk Container) AND at runtime; block deploys on critical findings without exception.

Operational checklist

  • Develop. Document the trigger, the owner and the cadence.
  • Scope: risk.

Evidence to collect

  • Code Repo Scan. Source-control evidence — branch protection, required reviewers, CI status, SCA / SAST reports.
  • Vapt Scan. vapt_scan evidence

Repository and security testing evidence should demonstrate controlled build, test, validation, and release practices for AI systems, including defect handling and risk-based quality gates. Focus: Development of AI system.

Common pitfalls

  • CI security checks set to warn-only; the merge button is faster than the security feedback.
  • Scans that hit only the perimeter; missing authenticated scans of OS, runtime libs and base images so most CVEs are invisible.

How it maps to other frameworks

Computed from shared tags, control component and key-operation subjects — handy when you are reconciling controls across an audit.

Related controls in ISO/IEC 42001:2023

References

Topics: secure developmentdevsecopsapplication securityvulnerability management