Cybernion
← Insights

ISO 27090

Is ISO 27090 Mandatory or Certifiable?

25 Oct 2026 · Updated 25 Oct 2026

No on both counts. ISO/IEC 27090:2026 is a guidance document, not a regulation, so nothing in Australian law requires it, and it has no certification scheme, so an organisation cannot be audited or certified against it. The certifiable AI standard is ISO 42001. ISO 27090 is the AI threat reference you use within your governance, not a credential you hold.

Is ISO 27090 mandatory?

No. No Australian law requires ISO 27090, and none is expected. It is published as guidance, which by nature describes good practice rather than imposing legal obligations. That matches the wider picture for AI standards here: even ISO 42001, the certifiable AI management system standard, is not mandatory in Australia, where the September 2024 Voluntary AI Safety Standard remains voluntary while mandatory guardrails for high risk AI are under consultation. ISO 27090 adds no legal duty of its own. Its pull is practical and commercial, not regulatory.

Can you get certified to ISO 27090?

No. There is no certificate, no audit, and no accredited scheme for ISO 27090, because it is guidance rather than a requirements standard. Requirements standards are written with auditable “shall” clauses; guidance standards are not, so there is nothing for a certification body to certify against. If you need a verifiable AI credential for a customer or tender, that is ISO 42001 certification, delivered by an accredited body on a three year cycle. A claim of being “ISO 27090 certified” is not possible and should be treated as a red flag. For the fuller comparison, see ISO 27090 vs ISO 42001.

If it is neither, why does it matter?

Because it is the clearest reference for AI specific threats, from a trusted source. ISO 27090 gives security, risk and procurement teams a vendor neutral list of the attacks that are specific to AI, data poisoning, model theft, prompt injection and the rest, with guidance on detection and mitigation across the life cycle. You use it to assess your own AI systems, to frame the questions you ask an AI vendor, and to feed the risk assessment inside an ISO 42001 or ISO 27001 management system. The value is in better decisions, not a logo for the website. With the ACSC 2025-26 threat report naming AI as the defining force in the threat environment, having a credible threat reference to work from is worth more than a certificate would be.

What should an Australian organisation do with it?

Use it as a reference, and pursue ISO 42001 if you need a credential. If you develop or deploy AI, read ISO 27090, inventory your AI use, and assess the relevant threats against your controls. If a customer or tender needs proof of responsible AI, build toward ISO 42001 certification and use ISO 27090 to strengthen the security part of that work. Cybernion helps Australian organisations do both through ISO 42001 readiness and broader virtual CISO support.

Frequently asked questions

Is ISO 27090 mandatory in Australia?

No. There is no law requiring it, and none is expected, because it is a guidance document rather than a regulation.

Can an organisation be certified to ISO 27090?

No. It has no certification scheme. The certifiable AI standard is ISO 42001, the AI management system standard.

So what is ISO 27090 for?

It is a credible, vendor neutral reference for the security threats specific to AI, used to assess and treat risk in your own systems and your suppliers.


Written by Gaurav Vikash, an ASD endorsed IRAP assessor and senior cyber security leader with 18 years of experience across Australia, the UK and Asia, including CISO and senior security leadership roles. He holds CISSP, CISA, CISM and CRISC and is an ISO 27001 Lead Auditor and ISO 42001 Lead Implementer, and speaks regularly at industry conferences.

If you want a clear read on where AI standards leave you, talk to us.

Sources:

  1. ISO/IEC 27090:2026, Cybersecurity, Artificial intelligence, Addressing security threats and compromises to AI systems, October 2026
  2. ISO/IEC 42001:2023, Information technology, Artificial intelligence, Management system, December 2023
  3. Voluntary AI Safety Standard, Department of Industry, Science and Resources, September 2024

Last updated: 25 October, 2026

Talk to us about your engagement

A scoped proposal within one business day.