Australia is preparing tighter controls on how artificial intelligence is used across government, workplaces and digital services as Canberra attempts to balance faster adoption with stronger public safeguards.
The Albanese government’s emerging framework will place greater scrutiny on automated government decision-making, with new rules expected to prioritise transparency, accuracy, fairness and human accountability.
The move reflects concerns that increasingly sophisticated AI systems could be embedded in public services without sufficient oversight, particularly when decisions affect access to payments, services or other entitlements.
Human oversight for government decisions
Attorney-General Michelle Rowland will lead work on rules governing automated decision-making within federal departments and agencies.
While the details are still being developed, the framework is expected to require agencies to clearly explain when automated systems are being used and ensure important decisions are not made without appropriate human review.
The issue carries particular weight in Australia following the Robodebt scandal, which demonstrated the risks of relying on flawed automated processes in government administration.
Canberra has framed the reforms as part of a broader effort to establish public trust in AI while still allowing agencies to use the technology to improve productivity and service delivery.
Digital platforms face greater responsibility
The government is also progressing a proposed digital duty of care, which would shift more responsibility onto technology companies to identify and prevent foreseeable harm.
Rather than responding only after problems emerge, digital platforms and AI developers could be required to build safety protections into their products from the outset.
Further privacy reforms are also planned, including stronger protections governing how personal information is collected, processed and used by automated systems. Workplace safety rules will form another part of the government’s AI agenda as businesses adopt tools capable of monitoring staff, allocating work and influencing employment decisions.
Data centres under the microscope
AI’s physical infrastructure is also attracting regulatory attention.
Australia already hosts more than 200 data centres, which the government estimates consume electricity equal to about 2% of the national grid. Demand could rise sharply as technology companies expand cloud computing and generative AI operations.
The government plans to introduce standards covering where major facilities can be built and how developers manage their impact on electricity networks, water supplies and surrounding communities.
Prime Minister Anthony Albanese said Australia’s approach would seek to attract AI investment without weakening employment conditions, dividing communities or damaging the environment. The proposed standards are intended to build on existing expectations for data centre and AI infrastructure developers.
National coordination
A new Office of AI within the Department of the Prime Minister and Cabinet will coordinate policy across government.
Its remit is expected to extend across economic development, infrastructure, safety, skills and regulation, creating a central point for managing a technology that currently cuts across numerous departments.
The initiative builds on the National AI Plan released in December 2025, which set three broad goals: capturing economic opportunities, spreading the benefits of adoption and keeping Australians safe. The plan also aims to develop domestic capability, attract investment and support AI adoption across industries and communities.
The challenge will be turning those principles into enforceable rules without slowing investment or leaving critical protections to voluntary compliance.
For Australia, the next stage of AI policy will therefore be less about declaring technological ambition and more about determining who remains accountable when automated systems get decisions wrong.