Albanese proposes new AI regulator and mandatory standards

On 10 November 2026, the Federal Government published a media statement detailing plans for a new “AI safety regulator” and mandatory obligations for AI developers under proposed Australian AI Standards legislation (Proposed AI Legislation).  The new “AI safety and security protections” would target “frontier AI companies” operating in Australia as part of as part of Albanese’s “AI in Australia’s Interests agenda”.  The AI regulator and new legal obligations remain proposals and will not take effect until legislation is passed.

What would the Proposed AI Legislation cover?

The proposed AI safety regulator would enforce compliance, with “significant penalties” for breaches.  Companies would be required to:

• maintain security controls throughout the software development and deployment stages;
• identify and manage serious risks associated with their models;
• permit independent model evaluations;
• promptly notify the Federal Government of incidents;
• maintain and publish safety and security frameworks; and
• comply with updated standards and rules as technology changes.

What don’t we know about the Proposed AI Legislation?

The announcement does not include any details about the proposed:

• penalty amounts;
• timelines for drafts of legislation; or
• criteria for determining which companies and AI models would be covered.

Why is the Government proposing this legislation?

The Proposed Legislation follows a review into an OpenAI agent’s unauthorised access to governmental online systems, including the Medicare Statistics Reporting Service Portal.  The Federal Government has accepted the review’s eighteen (18) recommendations and commenced implementation work.

The review also recommends considering criminal law reforms, including penalties for serious regulatory non-compliance and a standalone “failure to prevent” offence.  The latter would address circumstances where a company develops or implements an AI model that causes harm and fails to take reasonable steps to prevent it.  This is a recommendation to consider an offence; its elements and any defences have not been settled by the announcement.  It is to be seen whether any draft legislation for these AI safety proposals include this specific recommendation.

Takeaways about the proposed AI Regulator

The announcement is somewhat lacking in detail.  For example it does not say that every business using AI will fall within the new regime.  Any draft legislation will need to be examined carefully to determine its application to developers, software suppliers and businesses deploying third-party AI models.

Links and further references

Prime Minister of Australia, Media Release, New Australian AI safety and security protections.

Australian Federal Government, Rapid Review into Australian Government arrangements for an AI‑driven cyber incident recommendations.

Further information about AI Compliance

If you need advice on AI compliance or you are a software developer responsible for implementing AI systems in organisations and you want to know how to do this safely, contact us for a confidential and obligation free discussion.

Disclaimer

This article contains general commentary only. You should not rely on the commentary as legal advice. Specific legal advice should be obtained to ascertain how the law applies to your particular circumstances.

---

Related insights into AI