Government use of automated AI decision-making to be curbed under new Australian rules
The Australian government is preparing tough new rules to govern how federal departments and agencies use artificial intelligence in automated decision-making, as part of a broader national AI plan covering consumer protections, workplace safety and privacy. The move comes amid rapid growth in AI use and datacentre construction, with ministers seeking to build in fairness, accuracy and transparency after the illegal robodebt scheme exposed the dangers of unreliable automated processes replacing human judgement.
Attorney general Michelle Rowland will lead development of the government rules, while communications minister Anika Wells is progressing digital duty of care legislation requiring AI companies to build in safety features. A second round of privacy law reforms is also planned, alongside workplace AI safety measures. The plan follows prime minister Anthony Albanese's recent speech announcing a new office of AI and stronger protections for creatives, plus incoming rules for large datacentres from next year. Assistant minister Andrew Charlton noted Australia already has more than 200 datacentres, consuming around 2% of the energy grid, with capacity potentially tripling within a decade.
- Australia to tighten rules on government use of AI decision-making
- Digital duty of care law and privacy reforms also planned
- New standards coming for Australia's booming datacentre sector
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Supporters of tighter rules argue that automated systems making decisions about welfare payments, visas or benefits have repeatedly caused harm when deployed without adequate oversight, echoing failures such as Robodebt, and that citizens deserve a right to explanation, human review and redress when a government algorithm affects their life. They see this as a necessary safeguard for due process, transparency and public trust, ensuring efficiency gains from AI do not come at the cost of fairness or accountability, particularly for vulnerable people least able to challenge wrongful decisions.
The case against
Sceptics of heavy-handed curbs argue that excessive compliance requirements could slow or deter the adoption of AI tools that genuinely speed up processing times, reduce backlogs and cut costs for taxpayers, at a time when government agencies are already stretched. They contend that well-designed automated systems can be more consistent and less prone to individual bias than manual decision-making, and that onerous new rules risk entrenching inefficiency, adding red tape, and pushing agencies toward overly cautious, resource-intensive processes rather than encouraging responsible innovation.