ISO/IEC 27090 AI Cybersecurity Overview
AI Threat Detection and Mitigation Program
Speak to a GRC expert
If you have any questions or need assistance, please don't hesitate to contact us.
Share on social media
Purpose of ISO/IEC 27090: Addressing AI Security Threats
The standard helps organizations assess AI-specific compromises and select effective detection and mitigation measures.
Training Data Integrity
Safeguard datasets and data pipelines from poisoning, unauthorized changes and malicious contamination.
Model Security Governance
Assign clear ownership for model protection, AI risk decisions and security review activities.
AI Security Advisory Services
Apply structured threat assessments and practical safeguards informed by ISO/IEC 27090 guidance.
Threat Reduction Strategy
Reduce exposure to adversarial attacks, model theft and compromised outputs.
ISO/IEC 27090 Milestones:
The evolution of dedicated AI cybersecurity guidance, from standard development to the published 2026 edition.
Project Initiated 2022
International standards work began to address security threats unique to AI systems.
Final Approval 2026
The final draft progressed through international approval during 2026.
Published October 2026
ISO/IEC 27090:2026 was published on 14 October 2026 as AI security guidance.
Why ISO/IEC 27090 Matters: Protecting AI Assets and Trust
AI systems introduce distinct attack paths that require safeguards beyond conventional IT security.
Model Integrity
Protects models against manipulation that can undermine accuracy and trust.
Threat Resistance
Adds AI-specific defenses to established cyber and information security controls.
Security Alignment
Complements ISO/IEC 27001 and ISO/IEC 27002 without replacing them.
Stakeholder Trust
Demonstrates disciplined attention to AI threats, reliability and security assurance.
Risk Visibility
Highlights threats to data, models, outputs and critical AI-enabled decisions.
Business Resilience
Supports the secure deployment and operation of business-critical AI services.
Continue Reading
Ready to learn more about Applicability: AI Developers, Operators and Security Teams?