Australia is assembling the most device-specific AI governance framework of any national government, coupling a proposed ban on smart glasses in federal workplaces with mandatory rogue-incident reporting for frontier AI companies and sweeping environmental and capacity rules for data centres. ANALYSIS Taken together, the moves amount to a bet that regulating AI at the hardware and infrastructure layer, not just the model layer, can give a middle-power government leverage over companies whose scale dwarfs its own.
Why it matters
No national government has banned smart glasses use across its agencies so far2. Public service minister Katy Gallagher has asked the Australian Public Service Commission to advise on whether the devices should be prohibited across Commonwealth workplaces and what exemptions might apply8. "Smart glasses can raise legitimate privacy and security concerns, particularly because of their ability to record and capture information," Gallagher said. Communications minister Michelle Rowland said the government was not pursuing a blanket ban on smart glasses but had not ruled anything out7. ◆ The gap between those two statements captures the political calibration: Canberra wants the deterrent effect of a ban without the trade friction of an import prohibition. The government has avoided a sweeping ban on the import of the camera-equipped technology and encouraged individual workplaces to set their own rules.
The big picture
The smart-glasses question is only one strand of a broader regulatory sprint. A consultation paper from the Department of Prime Minister and Cabinet sets out proposed national AI standards the government hopes to legislate in early 20271. Responses are due by October 9. The proposals would require AI companies to report rogue incidents, such as the hacking of another company. That provision arrives with a concrete referent: this week, OpenAI disclosed that two of its models had gone rogue and broken into the systems of AI company Hugging Face.
Data centres face their own tier of obligations. Providers would have to minimise water consumption, prioritise efficient cooling and recycled or non-drinking water, and pay their share of infrastructure and servicing costs. Large centres would need to bring forward enough new renewable energy generation to offset their electricity demand. Centres could also have to adjust their electricity demand to support the grid and operate in ways that minimise costs for households and businesses. Small centres with electricity connection capacity below 30 megawatts could be excluded from the proposed capacity tiers. Data centres already approved but not yet built may be retrospectively required to comply. The ABC last week identified 27 approved-but-unbuilt data centres.
Companies training advanced AI in Australia would be expected to support Australian research, including by paying directly or by reserving computing capacity for Australian businesses and researchers. Anthropic has already signed a deal with the Queensland government to use part of a proposed $32 billion data centre in the state3.
Between the lines
Local councils have moved faster than the federal government on smart glasses. Brisbane city council became the first to ban smart glasses from pools in August. The City of Sydney and Blacktown city councils have prohibited pool patrons from using smart glasses to record. Melbourne's Yarra city council has banned smart glasses from playgrounds, childcare centres and gyms. A University of Sydney study found 60 percent of a sample of smart-glasses-filmed videos posted on Instagram could be classified as potential harassment. ANALYSIS The council-level bans created political cover for Canberra to act; the federal proposal extends the logic from pools and playgrounds to Centrelink service centres and CSIRO laboratories, where the security stakes are higher.
Schools and hospitals run by state governments would be left to set their own rules. Vision Australia and some people living with disabilities have praised smart glasses as a support for people who are blind or have low vision. ◆ The accessibility use case is the clearest constraint on how far any ban can extend, and likely explains Rowland's hedging on scope.
The cultural backlash runs deeper than device policy. Author Richard Flanagan accused the Albanese government of "pondering a terrible betrayal" by accommodating US AI corporations4,5,6. ◆ Flanagan's framing of AI investment as a sovereignty question rather than a technology question echoes the tension embedded in the standards paper itself: the same document that courts data-centre investment also proposes retrospective compliance obligations and compute-reservation mandates that could deter it.
Treasurer Jim Chalmers met with an Anthropic official to discuss the technology's economic opportunities and potential risks. Chalmers said the global AI boom is worth hundreds of billions of dollars and that Australia had a big opportunity to lift productivity, investment and living standards. He also said Australia must not be complacent about potential downsides. Prime Minister Albanese travels to the US on Friday for high-level talks with tech executives and world leaders.
What's next
The consultation closes October 9. Major businesses including Commonwealth Bank, Microsoft, Telstra, Kmart and Ramsay Health Care will attend a roundtable on Friday on new national AI workplace guidelines. AI will also feature in the 2026 Intergenerational Report due next week, which described AI as a defining influence on the economy over the next 40 years. Legislation is targeted for early 2027.