Regulatory guide
AI governance for aged care boards
Directors of aged care providers already carry legal governance duties for quality and safety. AI is now part of the operations those duties cover, and boards need a simple, regular way to oversee it.
Reviewed 15 November 2026

In short
AI governance for aged care boards does not need a new framework built from scratch.
Under the Aged Care Act 2024, the governing body of a provider in registration categories 4 to 6 must oversee all parts of the provider's operations, lead a culture of safety and quality, keep policies current and integrate clinical governance with corporate governance. AI tools fall inside those duties.
For how to govern AI itself, the current Australian Government guidance is the Guidance for AI Adoption, published by the National AI Centre on 21 October 2025, which sets out 6 essential practices and evolves the earlier Voluntary AI Safety Standard. Both are voluntary for providers.
In practice a board needs four things: a short AI policy, an AI register linked to the risk register, regular reporting from management, and the right questions at the table. Management runs the tools; the board makes sure someone is accountable for each one.
Board duties under the Aged Care Act 2024
The Aged Care Quality and Safety Commission says registered providers in categories 4 to 6 must make sure their governing body sets the strategic priorities for the organisation and promotes a culture of safety and quality (as at November 2026). Under the strengthened Quality Standards, the governing body must lead a culture of quality, safety and inclusion (Outcome 2.2), oversee all parts of the provider's operations, use a quality system and keep policies and procedures current (Outcome 2.3), and integrate clinical governance into corporate governance (Outcome 5.1).
Providers registered in Category 5 (Nursing and transition care) and Category 6 (Residential care) have further obligations.
Their governing body must have a majority of independent non-executive members and at least one member with clinical care experience, unless an exemption applies, for example for providers delivering funded aged care to fewer than 40 older people with a governing body of fewer than 5 members.
They must also have a quality care advisory body that gives the governing body a written report at least once every 6 months, and must offer to set up consumer advisory bodies at least once every 12 months. Board members are responsible persons, and the provider must assess their suitability at least once a year.
None of these duties mentions AI. They do mean that a board which cannot say what AI its organisation uses, or who is responsible for it, has a gap in its oversight of operations.
| Obligation | Who it applies to | What it means for AI |
|---|---|---|
| Oversee all operations; keep policies current (Outcome 2.3) | Categories 4 to 6 | Know what AI is used, approve an AI policy, review it |
| Culture of quality, safety and inclusion (Outcome 2.2) | Categories 4 to 6 | Staff can raise concerns about AI tools; AI should not exclude people who find technology hard |
| Clinical governance integrated (Outcome 5.1) | Providers delivering clinical care | Clinical governance committee reviews any AI that touches clinical records |
| Quality care advisory body reports at least every 6 months | Categories 5 and 6 | Ask the advisory body to comment on AI use affecting quality |
| Consumer advisory bodies offered at least every 12 months | Categories 5 and 6 | Ask residents and families how automated contact is working for them |
| Suitability of responsible persons assessed yearly | All registered providers | Board AI literacy is part of being able to oversee operations |
The Australian Government's AI guidance
The Department of Industry, Science and Resources states that on 21 October 2025 it published the Guidance for AI Adoption, through the National Artificial Intelligence Centre, which outlines 6 essential practices for safe and responsible AI governance and "evolves" the Voluntary AI Safety Standard (as at November 2026). The Voluntary AI Safety Standard, first published on 5 September 2024 with 10 voluntary guardrails, is still available. The new guidance comes in a foundations version for organisations starting out and an implementation version for higher-risk or more complex use.
The 6 essential practices are: decide who is accountable; understand impacts and plan accordingly; measure and manage risks; share essential information; test and monitor; and maintain human control. The guidance also recommends keeping an AI register, making supply chain accountability clear, and maintaining alternative pathways so critical functions can continue if an AI system fails or is taken offline. That last point matters in aged care: if an AI phone agent stops working at 2am, families still need to reach the on-call nurse.
The guidance is voluntary. For an aged care board it is a sensible reference because it lines up with duties directors already have under the Act: accountability, risk management, information to stakeholders and human oversight.
What an aged care AI policy should cover
An AI policy for an aged care provider can be two to four pages. Its job is to set boundaries that managers can apply day to day without coming back to the board. A workable outline:
- Purpose and scope: which tools, services and staff the policy covers, including contractors and volunteers
- Accountability: the executive who owns AI overall, and a named owner for each tool
- Permitted uses: drafting, summarising, routing, reminding, answering approved questions
- Prohibited uses: clinical diagnosis, medication decisions, care decisions, admission decisions, and entering resident or client information into unapproved public tools
- Human review: which outputs a qualified person must check before they are used or sent
- Privacy and data: data location, retention, vendor training on your data, overseas disclosure and privacy impact assessment before go-live
- Transparency: how older people, families and staff are told when they are dealing with AI, and how to reach a person
- Incidents and complaints: AI errors go through the incident management system and complaints process
- Approval and review: who approves a new AI tool, and an annual policy review
Apply this to your service
Talk it through with a Yes AI consultant in a free 30-minute call, including whether AI fits your service at all.
An AI risk register that fits your existing one
A separate AI risk system is rarely needed. Add an AI section to the existing risk register, with one line for each tool, and link it to an AI register that records what each tool does, what data it touches, which vendor runs it and who owns it. Rate each risk with the same likelihood and consequence scale you use elsewhere so the board can compare it with other risks.
| Risk | Example | Controls |
|---|---|---|
| Inaccurate output | AI-drafted progress note leaves out a fall | RN reviews every draft; incident process for errors; periodic audit sample |
| Privacy breach | Staff paste resident details into a public chatbot | Approved tools only; training; access logs; breach response plan names AI |
| Over-reliance | Staff stop checking AI summaries after a few weeks | Spot checks; review rates reported monthly; refresher training |
| Service outage | AI phone agent unavailable overnight | Fallback call routing to on-call staff; outage tested twice a year |
| Vendor change | Vendor changes model, terms or data location | Contract notice clauses; annual vendor review; exit and data return plan |
| Exclusion | Older people with hearing loss struggle with the phone agent | Easy route to a person; feedback from consumer advisory body |
Reporting AI to the board
Boards should not need a separate AI meeting. Fold AI into the reporting already going to the board and its quality and risk committees. A one-page quarterly report usually covers it: the AI register with any tools added or retired, incidents and complaints involving AI, results of output audits (for example, the share of AI-drafted notes that needed significant correction), privacy matters, and any changes in law or guidance.
Ask management to report in plain numbers rather than adjectives. "We audited 50 AI-drafted progress notes and 3 needed a clinical correction" tells the board far more than "the tool is working well". If the quality care advisory body or consumer advisory body has raised AI, include that too.
Questions directors should ask
Directors do not need to understand how AI models work. They need to be able to test whether management has control. These questions work at most board tables.
- What AI tools do we use, including any staff have adopted on their own?
- Who is accountable for each tool, and who checks its outputs?
- Where is resident and client data processed and stored, and is any of it used to train the vendor's models?
- Which decisions does AI influence, and can a person always overturn the result?
- How are older people and families told they are dealing with AI, and how quickly can they reach a person?
- What happens if the tool fails tonight?
- What AI incidents or complaints have we had, and what changed as a result?
- Does our privacy policy need updating for the automated decision rules that start on 10 December 2026?
How it works in practice
- 1
Brief the board
A short session on what AI the organisation uses now and the governance duties that apply to it.
- 2
Approve an AI policy
Adopt a two to four page policy using the outline above, owned by a named executive.
- 3
Build the AI register
Management lists each tool, its owner, data and vendor, and links it to the risk register.
- 4
Set reporting
Add a one-page AI section to quarterly quality and risk reporting.
- 5
Review yearly
Review the policy, the register and the government guidance each year, or sooner when the law changes.
Common questions
Do aged care boards have legal duties about AI?
Not AI-specific duties. But governing bodies of providers in categories 4 to 6 must oversee all operations, keep policies current and lead a culture of safety and quality, and AI tools fall within those duties.
What is the Guidance for AI Adoption?
It is the Australian Government's current voluntary guidance on AI governance, published by the National AI Centre on 21 October 2025. It sets out 6 essential practices and evolves the Voluntary AI Safety Standard.
Is the Voluntary AI Safety Standard still relevant?
It is still published, with its 10 voluntary guardrails, but the government describes the Guidance for AI Adoption as the updated and simplified guidance that evolves it. Start with the newer guidance.
Does an aged care provider need an AI policy?
There is no specific legal requirement, but a short AI policy is the simplest way to show the governing body oversees AI as part of operations and keeps policies current, as the Quality Standards require.
Should the board own AI, or management?
Management. A named executive should own AI overall, with a named manager for each tool. The board oversees that arrangement, and management runs the tools and reports on them to the board.
Do board members need AI training?
Enough to ask good questions. The Guidance for AI Adoption recommends AI literacy training for people with accountability, and directors are responsible persons whose role includes overseeing operations.
Sources and further reading
- Provider governance obligations (Aged Care Quality and Safety Commission)
- About workforce obligations (governance and Quality Standards outcomes) (Aged Care Quality and Safety Commission)
- Voluntary AI Safety Standard (and Guidance for AI Adoption notice) (Department of Industry, Science and Resources)
- Guidance for AI adoption: implementation guidance (National Artificial Intelligence Centre)
- New resources on transparency for use of AI and automated decision-making (Office of the Australian Information Commissioner)
Related pages
Aged Care Act 2024 and AI
The new Act changed who is accountable for what in aged care. It does not mention AI tools, but every duty it places on providers and workers still applies when part of the work is done with AI.
AI Risks and Limits in Aged Care
We build AI tools for aged care providers, and we think the risks deserve as much attention as the benefits. This page sets out what can go wrong, how to reduce it, and the jobs AI should not be given.
AI Privacy Guide for Aged Care
Aged care records hold some of the most sensitive information a person has. Before any AI tool touches it, you need to know where the information goes, who can see it and what the law expects.
Multi-Site Aged Care Groups
A group with five homes usually has five slightly different ways of answering the phone, logging an enquiry and chasing a family. AI tools set up once at group level give every site the same process, while each home keeps its own staff, contacts and local detail.
Talk it through with us
A free 30-minute video call about where admin time goes in your service and which task is worth automating first. No obligation, and no resident information needed.