Australia Has Six AI Inquiries and Zero AI Policy. Here’s What That Means for Your Business.
Last week, journalist Cameron Wilson had to count Australia’s AI inquiries on X because no one in government was keeping track. The final tally: six. Three federal processes, three state-level, all running simultaneously, none of them talking to each other.
Six inquiries. Zero AI policy.
That gap between inquiry and action is where Australian businesses are currently operating.
What Each Inquiry Is Actually Looking At
The Joint Select Committee on Artificial Intelligence is the broadest. It covers economic opportunity, workforce disruption, national security, copyright, sovereign AI capability, and fraud risks. Bipartisan origin. Reports 30 November.
The Senate Environment and Communications Committee is examining a narrower slice: data centres, energy consumption, water usage, community impacts. The committee held hearings in Sydney on 22 September and Melbourne the day after. Reports 16 November.
Then there’s the one most coverage has missed entirely. The PM&C “Getting It Right” consultation, run by the Office of AI inside the Department of Prime Minister and Cabinet, is open right now, accepting submissions until 9 October. Unlike the two parliamentary committees, this one feeds directly into legislation expected in early 2027. It’s the most consequential of the three federal processes and the least discussed.
At state level: NSW wrapped hearings in May and reports by 30 September. South Australia’s Royal Commission, yes, a Royal Commission, the heavyweight instrument, start 1 October and reports by July 2027. Tasmania just closed submissions on its own data centre inquiry. That’s six bodies examining six slices of the same problem.
Energy. Ethics. Data centres. Workforce. Societal impact. Economic opportunity. Each committee has its lane. Nobody is steering the whole car.
The States That Skipped the Inquiry Altogether
Victoria and Queensland tell you more about Australian AI governance than any of the six inquiries do.
Victoria, which has more than 50 operating data centres and 30-plus in the pipeline, looked at the inquiry model and decided it was too slow. On 22 September 2026, the Victorian government announced new rules: data centres banned from residential zones, mandatory 150-metre buffer zones, renewable energy requirements, no drinking water. Announced, not legislated, the rules don’t apply retroactively, which critics have called a significant gap. But Victoria moved. No inquiry. Direct policy.
Queensland went the other direction. Premier Crisafulli successfully lobbied for a carve-out at National Cabinet, allowing Queensland to power AI data centres with coal and gas. Anthropic, the company behind Claude, announced a $32 billion data centre deal in Queensland shortly after. An inquiry would have complicated that. So there’s no inquiry.
Two states, two completely different calculations. Neither ran an inquiry.
What the Rest of the World Did
Singapore built practical national AI guidelines in months and updated them as the technology changed. Not thorough, not perfect, but deployed.
The EU went thorough with the AI Act, the most detailed framework anywhere, and took years. It’s still being implemented. The act was finalised in 2024; full compliance requirements don’t kick in until 2026-2027.
The UK chose sector-by-sector guidance rather than blanket legislation, which is faster but creates uneven coverage depending on your industry.
Nobody has landed on the obviously correct model. But all three approaches involved decisions, not just inquiries.
Australia has the inquiries. The decisions are scheduled for late 2026 to 2027.
What This Means If You Run a Business
While the committees are hearing witnesses, businesses are making AI decisions. Your competitors aren’t waiting for November’s reports. They’re deploying tools, rewriting workflows, and cutting costs. The businesses adopting AI now will have six to twelve months of operational experience by the time any federal legislation arrives. That lead compounds.
Waiting for Canberra is a strategy, but it’s a losing one. AI regulation in Australia will eventually arrive, probably in 2027, probably focused on larger enterprises and data centre operators first. The things that matter to SMBs, the day-to-day stuff around customer data, AI-generated content, automated decisions, those will be addressed last, if at all.
The only AI policy you can control right now is your own. Decide what data you feed into AI tools. Decide how you disclose AI use to customers. Decide what decisions need a human in the loop. Write it down. Update it when the tools change.
Six inquiries can’t do that for you. Only you can.
📖 More deep dives: News Deep Dives
📊 Compare AI tools side by side | 💼 Free resources & AI prompt packs
Sources
- Joint Select Committee on Artificial Intelligence
- PM&C “Getting It Right” consultation
- Announced, not legislated, the rules don’t apply retroactively
- count Australia’s AI inquiries on X
More hot takes: Hot Takes →
If this interests you
- Albanese Says Keep Humans in Control of AI — But Are We Actually Doing That?
- Is Your Australian Online Shop Ready for AI Buyers? (Perplexity Just Added a Buy Button)
- The AI That Wrote Its Own Jailbreak
More hot takes: Hot Takes →
