Australia needs a human‑rights act as AI tools reshape aged‑care decisions
Australia is racing to embed artificial‑intelligence systems across government services while a federal human‑rights framework remains absent. A parliamentary committee, a massive public petition and state‑level legislation have highlighted the gap, but the national parliament has yet to pass comprehensive rights protections. The government’s “digital duty of care” promises limits on algorithmic…
Key points
- Aged‑care’s automated integrated assessment tool (IAT) is failing, denying timely services to vulnerable seniors.
- Australia lacks a federal human‑rights act, leaving AI‑driven decisions largely unregulated despite public support.
- Government mandates chief AI officers and AI‑use disclosures, but privacy statements remain unread and lack human‑review rights.
The most urgent example is the Department of Ageing’s Integrated Assessment Tool (IAT), an AI‑driven system that decides eligibility for aged‑care services. Within months of deployment, health officials reported widespread mis‑classifications, leaving vulnerable seniors without needed care and pushing hospitals to capacity. Calls for human oversight, mandatory chief AI officers, and updated privacy disclosures have so far failed to provide a clear path for review or redress, exposing a legal vacuum that could erode basic rights.
Advocates such as MP Kate Chaney and former aged‑care inspector Natalie Siegel‑Brown argue that without enforceable rights, algorithmic decisions risk repeating past scandals like the Robodebt saga. The article warns that the lack of a federal human‑rights act could allow opaque AI systems to dictate essential services, making legislative reform a pressing priority for both citizens and policymakers.
As Australia faces an AI-generated future, a human rights act is needed more than ever
The Guardian AI · 12 September 2026
As we hurtle towards a post-human future, the absence of a human rights act in Australia will put us all at ever greater risk.
Human rights, which the Australian parliament has never managed to properly legislate despite overwhelming public support, risks becoming a dimly remembered artefact of another age at precisely the moment we need it most.
Artificial – authoritarian – intelligence, automated decision-making, independent assessment tools and algorithms will shape the lives we live and the public support we receive.
But human rights are not adequately legislated in Canberra – despite the recommendation of a parliamentary committee, a huge petition and campaign and effective laws in Victoria, Queensland and ACT. Now rights can easily be lost in an environment shaped by opaque algorithms and computers that say no.
The government has been patting itself on the back with its proposed digital duty of care – “my feed my way”, as the prime minister announced with uncharacteristic pithiness. It aims to introduce limits on the algorithms perfected by the global surveillance giants and give Australians the right to help shape the information they receive (but not its accuracy).
Meanwhile, the government is racing to use similar tools to that determine how much support people receive for essential services when they need them most. Kate Chaney’s private member’s bill is the only sign of pushback.
Hundreds of thousands of public servants have been trained on how to use artificial intelligence. Yet the information commissioner found that few departments declared their use of AI.
GovAI is set to become something more than an announcement. Every department and agency was required to appoint a chief AI officer by July.
Come December, the privacy statements of every department – and company – that uses AI to make decisions that impact human beings will have to be updated to note this.
But who reads privacy statements before clicking consent?
The law falls short of the robodebt royal commission recommendations. It does not require companies and departments to reveal how the process works, the business rules and algorithms, or provide a right to human review of automated decisions.
The provision of care to older Australians is currently the biggest canary in this mine.
After only 10 months, it is clear that the department of ageing’s automated integrated assessment tool (IAT) – used to determine eligibility and access to services – is failing. Contrary to all best practise recommendations or the scarifying experience of robodebt, it has no human oversight or override options.
As Melissa Davey reported, within days of the tool’s adoption, alarm bells began to ring.
State health officials recognised that it was potentially catastrophic; spreadsheets of inappropriate decisions were compiled in every state. “Misaligned IAT outcomes”, they were called in NSW. In the words of the head of WA Health: “Without a pathway for discretion, a small but vulnerable group of older people might experience catastrophic outcomes while waiting for algorithmically determined access to services.”
Beyond the human pain and suffering, this matters a great deal for the states and territories. It is their hospital beds that become the only option for elderly people who cannot get adequate care at home, or access to a residential care bed.
The equivalent of whole hospitals are filled with people who should be receiving more appropriate care elsewhere.
For years we watched and listened to the horror stories of aged care as they were revealed to the royal commission. Before the commission finally reported, the pandemic exacerbated the problem as residential facilities became “no-go zones”, leaving the elderly marooned, their loved ones tapping on windows and staff fearful.
The human rights of thousands of people were gravely infringed.
Those designing the new Aged Care Act, which came into effect last November, took this seriously. The beginning of the act is a testament to the fundamental right to be treated with dignity, respect and independence.
In residential care homes all over Australia, the statement of the human rights of the elder person was tacked up on notice boards for those who could still read – and their loved ones – to note.
What was not so prominent was the inability of those residents to take legal action if their rights were breached. There was a complaints mechanism, but these were not fully legally enforceable rights.
The human right of the aged to care were again to be rationed – by algorithms, queues for assessments, caps on the number of places and access to providers, by regulations that made delivering services at home or in facilities onerous.
At the time, it was only the super-observant departmental officials in the states and some advocacy groups who realised access was also being rationed by an algorithm that could not be overridden by a human being.
The contrast between the frontend of the Aged Care Act and its backend implementation could not have been sharper. The beginning of the act describes the rights of older Australians to receive care and be treated with dignity and respect. But the implementation is about transactions: minutes of care, checklists of tasks, organisational obligations.
“The plumbing of the reforms is undermining the poetry of the act,” the former inspector general of aged care Natalie Siegel-Brown told the National Press Club.
Instead of life, it is the human rights of the elderly, that to paraphrase TS Eliot, are being “measured out … with coffee spoons”.
This text was published by The Guardian AI and written by Julianne Schultz. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us.
More in Policy & Regulation
All →- Meta hit with class action over secret AI face recognition tool NameTag · 5 src
- Greece Sets AI Rules for Schools · 1 src
- Senators probe OpenAI over Hugging Face breach and AI safety risks · 3 src
- Former UK PM Rishi Sunak Becomes Senior Adviser to Microsoft and Anthropic · 1 src
- US agencies accuse six Chinese AI firms of large‑scale model theft · 2 src
Comments
via GitHub Discussions