Evinact Director Jane Brimacombe explores what Australia’s proposed mandatory AI standards could signal for the future of regulation, and why organisations should be strengthening their AI governance now.
Australia is taking another step towards a more regulated AI environment.
On 10 November 2026, the Australian Government announced proposed mandatory AI Standards for companies developing or operating ‘frontier AI’, the most advanced AI models at the leading edge of current technological capability.
The proposal includes a dedicated AI safety regulator with powers to enforce compliance and impose significant penalties. The proposed requirements cover security throughout the AI lifecycle, identifying and managing serious model risks, independent evaluation, rapid incident reporting and transparent safety and security frameworks.
These are important protections as AI systems become more capable and autonomous. But the significance of this announcement may extend well beyond the companies initially in scope.
It raises a broader question for organisations adopting AI: are they prepared for a regulatory environment that may become increasingly demanding?
A turning point for AI regulation
Australia’s existing approach to AI safety has relied heavily on voluntary guidance, so the proposed mandatory standards represent a significant shift.
A dedicated regulator, enforceable requirements and standards that can evolve alongside technology suggest the Australian Government may be laying the foundations for a broader AI regulatory framework.
The Government has also indicated the legislation will address other AI policy areas, with further reforms under consideration.
The eventual scope and timing remain uncertain, and it would be premature to assume the proposed standards will automatically extend to all AI developers or users.
However, organisations shouldn’t assume today’s regulatory boundaries will define tomorrow’s obligations.
AI risks extend beyond the developers
For most organisations, the immediate challenge isn’t developing frontier AI models. It’s adopting and using AI developed by others.
AI capabilities are increasingly embedded in everyday business applications, cloud platforms and third-party services. They may be introduced through software updates, enabled within existing products, configured by business teams or connected to organisational data and processes.
The risks extend beyond the underlying AI model. They depend on how the technology is implemented, what information it can access, which decisions it influences and what controls are in place.
An AI model or service may be tested and secured by its developer, but that doesn’t guarantee it will be used appropriately in a particular organisational context. Nor does it establish who’s accountable when an AI-enabled process produces an incorrect, unfair or harmful outcome.
Preparing for what comes next
Australia’s existing voluntary Guidance for AI Adoption provides a foundation for responsible adoption. The proposed mandatory standards add another layer of accountability for frontier AI developers.
But organisations using AI shouldn’t wait for further regulation to strengthen their own governance.
Some practical questions worth asking now include:
- Visibility: Where is AI being used across the organisation, including within third-party products and services?
- Accountability: Who owns each AI-enabled solution, and who’s responsible for decisions made using its outputs?
- Data and information management: What information can AI systems access, how is it handled, and have the associated privacy, security and information risks been assessed?
- Human oversight: Can people meaningfully review, challenge or intervene in AI-enabled decisions?
- Ongoing assurance: How are systems monitored as their capabilities, applications and risks change?
These considerations are particularly important when AI arrives through existing software rather than a dedicated project, potentially bypassing the scrutiny applied to major technology investments.
Establishing a clear view of AI use, strengthening accountability and reviewing existing governance and risk management practices will help organisations manage today’s risks while preparing for potential future obligations.
The question is no longer simply whether an organisation is covered by today’s requirements. It’s whether it’s ready for what comes next.
To discuss how your organisation can strengthen AI governance and prepare for an evolving regulatory environment, contact Evinact.




