The Human Technology Institute at the University of Technology Sydney (UTS) published its AI Governance Snapshot #6 in May 2026. A 2025 survey of 419 Australian corporate leaders by the UTS Human Technology Institute found only about 15% felt confident in their organisation's AI governance.
For a definition first, read what AI governance is.
Define the scope
Identify the technologies and uses covered. Include relevant vendor features, models, agents and automated workflows. Different obligations use different definitions.
An AI inventory lists AI uses, and a disclosure map lists automated decisions that may need disclosing. The lists overlap, and each answers a different question. A rules-based decision can fall under privacy disclosure rules even if no one calls it AI.
Identify the use and owner
Record purpose, workflow, affected people, information, systems and accountable business owner. Distinguish confirmed facts from initial declarations.
The owner should understand what the use changes in operation.
Assess the consequence
Examine plausible benefits, failures and harms in context. Consider quality, privacy, security, fairness, resilience, vendor dependence and autonomy.
Bring in specialists where needed. Report legal issues and unacceptable failure modes separately from any overall score.
Decide what may proceed
Define the approval route, evidence, restrictions and deployment conditions. Set one standard for experiments and a different one for consequential production use.
Record the limits. Approval for one dataset or action does not cover every future use of the product.
Put controls into the workflow
Implement boundaries in access, interfaces, operating procedures and review. Make the human role specific.
Use technical and operational evidence to test that the controls work in operation, including the human role.
Monitor, respond and change
Track performance, incidents, complaints, exceptions and material changes. Identify who receives alerts, who can interrupt the system and how work continues if it becomes unavailable.
In the same UTS Human Technology Institute survey, 21.7% of organisations said they periodically or continuously monitor the performance of their AI systems.
Reassess changes to model, data, tools, permissions, purpose or affected population. When a use ends, remove its access and deal properly with its information.
Map the framework to the Guidance for AI Adoption
The National AI Centre published the Guidance for AI Adoption on 21 October 2025 as the first update to Australia's Voluntary AI Safety Standard. It sets out 6 essential practices. The table maps each practice to the steps on this page. The steps and the mapping are Applied AI Australia's own work. They are not a checklist set by law.
| Essential practice | Where it sits in this framework |
|---|---|
| Decide who is accountable | Identify the use and owner |
| Understand impacts and plan accordingly | Assess the consequence |
| Measure and manage risks | Assess the consequence; Decide what may proceed |
| Share essential information | Identify the use and owner; Decide what may proceed, where the limits are recorded |
| Test and monitor | Put controls into the workflow; Monitor, respond and change |
| Maintain human control | Put controls into the workflow, where the human role is made specific and tested |
Integrate with existing governance
Finance, procurement, privacy, security, risk and internal audit already make decisions AI touches. Connect AI decisions to those existing processes.
The AI Risk Management Framework from the US National Institute of Standards and Technology is a voluntary reference. It does not certify an organisation or replace Australian law.
Test the framework on one actual use
Follow it from purpose to approval, controls, monitoring and response. Where evidence is missing, assign an action and an owner to close the gap.
The AI Roadmap applies this framework to your own priorities.
Frequently asked questions
What should an AI governance framework include?
It should take each AI use from approval to retirement, with a named owner at every step. Each step should leave a record: the inventory entry, the assessment, the approval decision and the monitoring reports.
What are the 6 essential practices in the Guidance for AI Adoption?
They are: decide who is accountable; understand impacts and plan accordingly; measure and manage risks; share essential information; test and monitor; and maintain human control. Both versions of the guidance cover all 6. Implementation guidance adds stronger controls for complex or higher-risk use.
Does guidance replace the law?
Voluntary guidance does not replace any law. The National AI Plan states that 'established laws remain the foundation for addressing and mitigating AI-related risks'.
How many Australian organisations monitor their AI systems?
Only a minority say they do, according to the survey figure under 'Monitor, respond and change'. Useful measures to start with are errors found in review, complaints and how often staff override the system.
Bring one AI use and the governance question your board needs answered.