Australia’s newest regulator does not have the luxury of a quiet start.
On 12 June 2026, the Australian Naval Nuclear Power Safety Regulator (ANNPSR) issued its first substantive licence – to Australian Naval Infrastructure (ANI), authorising site preparation for a facility to construct AUKUS submarines within the Osborne designated zone in South Australia.
It was, on paper, a routine regulatory decision. In practice, it was the first real test of a regulator that had existed for barely seven months, built almost entirely from nothing, for an activity that has never before occurred on Australian soil.
Two months earlier, in April, ANNPSR’s Assistant Director-General for Licensing and Regulation, Paul Dailey, stood in front of members of the regulated community – engineers, lawyers, compliance officers from the defence and nuclear supply chains – and explained, in effect, how the rules worked.
“Now that ANNPSR has been established as a statutory regulator and is operating,” he told them, “it is important for us that those who will work with us from across the regulated community are given opportunities to be introduced to the ANNPS regulatory framework and its key features; the principles and systems, licensing and compliance and response to a nuclear safety incident.”
It’s an unusual sentence for a regulator to say out loud. A body already issuing licences was still introducing itself to the people it licenses. A regulator built and operated at the same time, for an activity built at the same time, is a fair description of ANNPSR’s first year. It’s also worth taking seriously as a case study, not just a curiosity.
This is the story of what that looks like in practice, and what it might mean for regulators asked to do something similar in future.
Nothing to inherit
Most regulators inherit something. A body of law that has been tested in courts and tribunals. A cohort of licensees who understand, even if they resent, what is expected of them. A professional community with shared norms and a rough consensus on where the lines sit. Even a poor inheritance is still an inheritance – it gives a regulator somewhere to start.
ANNPSR started with almost none of that. Australia had never regulated naval nuclear propulsion before. As the regulator’s own frequently asked questions page puts it, the country’s existing nuclear regulatory framework “was not designed to accommodate the unique activities associated with naval nuclear power in a military context.” The Australian Radiation Protection and Nuclear Safety Agency (ARPANSA), which had handled a handful of related licences until ANNPSR’s establishment, was built for civilian activities – medical isotopes, research reactors, industrial radiography – not for a military submarine construction and sustainment programme.
ANNPSR was established under the Australian Naval Nuclear Power Safety Act 2024 (Cth), which came into force on 1 November 2025 alongside the Australian Naval Nuclear Power Safety (Transitional Provisions) Act 2024 and the Australian Naval Nuclear Power Safety Regulations 2025. ARPANSA’s relevant licences, tied to HMAS Stirling in Western Australia, transferred automatically that day. Everything else – the Osborne designated zone, the licensing processes, the compliance culture, the regulated community itself – had to be built from that starting point.
Australia has no domestic civil nuclear power industry. There is no pool of experienced nuclear engineers, reactor operators or safety inspectors to draw on.
Compounding the problem: Australia has no domestic civil nuclear power industry. There is no pool of experienced nuclear engineers, reactor operators or safety inspectors to draw on. ANNPSR’s own corporate plan names this squarely as a risk: that the regulator “fails to recruit, retain and support a skilled nuclear workforce in the context of domestic and global shortages of nuclear skilled, qualified and experienced persons.”
Keith Manch, one of the architects of the regulatory profession’s development in New Zealand, has described an “80/20 principle” that runs through good regulatory practice: regulation, he argues, is roughly 80 per cent the same across every domain, and 20 per cent domain-specific. It is a useful frame, and one worth testing against ANNPSR. The problem for a genuinely new regulator is that it must apply the 80 per cent – consultation, proportionality, evidence-based decision-making, transparency – before it has any real sense of what the 20 per cent looks like in practice. It is applying a template while still discovering the shape of what the template needs to fit.
Built while operating
ANNPSR did not have the option of a long design phase followed by a clean launch. The Australian Government wanted a functioning regulator in place as the AUKUS submarine programme moved from planning into construction, and it wanted that regulator operating independently from day one.
The result is a body that has been making consequential decisions – granting licences, setting compliance expectations, defining designated zones – while still building the institution that makes those decisions. The corporate plan is candid about the trajectory: ANNPSR’s “regulatory role will significantly evolve in scope and complexity over the next 10 years, commensurate with the execution of the Optimal Pathway,” the staged plan under which Australia will host rotational visits by UK and US nuclear submarines from as early as 2027, acquire a small number of second-hand US Virginia class submarines through the 2030s, and eventually build SSN-AUKUS submarines domestically at Osborne.
The regulator launched with more than 70 trained staff and a headquarters in Canberra. It is aiming for a workforce of roughly 125 people by 2031–32, a size calibrated to a regulated population that barely exists yet. In the meantime, ANNPSR has leaned heavily on its AUKUS partners: personnel are embedded with counterpart agencies in the United States and the United Kingdom, and in February 2026 the US Naval Reactors programme – whose military nuclear propulsion record runs to more than 70 years without a reactor accident – visited ANNPSR’s Canberra headquarters. In July 2025, before ANNPSR formally existed, its predecessor design team travelled to the United Kingdom to meet the Office for Nuclear Regulation (ONR) and its counterpart military regulator, the Defence Nuclear Safety Regulator (DNSR), visiting Rolls-Royce’s submarine reactor plant in Derby and the naval bases at Devonport and Barrow.
Mike Finnerty, ONR’s Chief Executive and Chief Nuclear Inspector, described the exchange in terms that say as much about what Australia was missing as what it was gaining. “One of the objectives of the meeting was to share information about our respective licensing regimes and provide the delegation with an understanding of ONR’s regulatory approach to nuclear submarine construction and maintenance, and off-site emergency planning,” he said. “The discussions were extremely constructive, and we identified a number of ways in which we can continue our collaboration for the benefit of our two nations.”
“The discussions were extremely constructive, and we identified a number of ways in which we can continue our collaboration for the benefit of our two nations.”
– Mike Finnerty, Chief Executive and Chief Nuclear Inspector,
Office for Nuclear Regulation (UK)
That exchange is regulatory capital being transferred between institutions – decades of accumulated licensing experience, safety culture and incident response practice, handed to a regulator that does not yet have any of its own. Building that capability ahead of the workload, rather than scrambling to build it once problems have already occurred, is rare in any jurisdiction, and it is the clearest sign so far that ANNPSR’s early design choices have been deliberate rather than reactive.
For regulators elsewhere, the lesson generalises. Technology decisions, staffing models and case management systems adopted in a regulator’s first 12 months tend to calcify. Choices made for a small, specialised, relationship-heavy caseload can prove badly suited once the regulated population expands and diversifies – exactly the trajectory ANNPSR’s own corporate plan predicts for the next decade.
A very small regulated community
Regulatory capture is usually described as a slow process: a large, well-resourced industry gradually shaping the rules that are meant to constrain it. ANNPSR faces a different, more structural version of the same risk. Its regulated community, for now, consists of a handful of entities – ANI at Osborne, the Australian Submarine Agency, a small number of contractors supporting HMAS Stirling. All of them are pursuing the same national programme as ANNPSR itself. All report, ultimately, to the same minister.
Marie Doole, whose work on regulatory capture is well known to TMR readers, has argued that capture often begins not with corruption but with drift – a slow alignment of interests between a regulator and the entities it oversees, particularly where the regulator’s own success is bound up with the programme’s success. That is precisely ANNPSR’s structural position. A failure at Osborne or Stirling would not just be ANNPSR’s failure; it would be a setback for the entire AUKUS submarine programme, and by extension, a strategic and diplomatic problem for the Australian Government.
The ANNPS Act tries to manage this with formal safeguards. Former Defence staff members – including Australian Defence Force personnel and Department of Defence or Australian Submarine Agency employees – must wait at least 12 months after leaving those roles before joining ANNPSR. The regulator cannot be directed by Defence or the Australian Submarine Agency in the performance of its functions; the only exception is a ministerial direction issued in the interests of national security or in response to an emergency, and any such direction must be tabled in both Houses of Parliament. As the regulator’s frequently asked questions page states plainly: “The Regulator has complete discretion in performing its functions under the Australian Naval Nuclear Power Safety Act 2024 and is not subject to direction by any person in relation to the performance of those functions.”
Structural independence is not the same as cultural independence, though, and the ANI licence decision in June was the first real occasion for outside observers to test whether the two align. The site preparation licence covers construction of the facility where AUKUS submarines will eventually be built – a decision made under close public and political attention, in a state government area that had previously passed a council motion opposing storage of radioactive waste at Osborne. That the licence was granted, on schedule, without major public controversy, suggests the process held. Whether it continues to hold as the volume and complexity of licensing decisions grows is the open question.
The workforce Australia does not have
Dale Heinken’s appointment as ANNPSR’s Deputy Director-General, announced in April 2026, illustrates the practical shape of the talent problem. Heinken is an American who spent 26 years in the US Navy’s nuclear propulsion programme, commanding two warships and serving as a reactor officer aboard the aircraft carrier USS Theodore Roosevelt, before moving into senior roles in Australian defence industry. He is, in effect, imported expertise – a way of closing a capability gap that Australia’s own workforce cannot yet fill.
Director-General Michael Drake’s background is a different, more distinctly Australian version of the same improvisation. Drake spent 15 years in the Royal Australian Navy, largely aboard Collins class submarines, before retraining as a lawyer and moving through senior roles at NSW Government transport agencies and, most recently, as Executive Director of Operations at the Australian Maritime Safety Authority. He has been both inside a submarine and inside a regulator – a combination that barely existed as a career path in Australia before AUKUS made it necessary.
“Robust regulatory frameworks are critical to the success of Australia’s nuclear-powered submarine programme.”
– Dr Gillian Hirth,
CEO, ARPANSA
Dr Gillian Hirth, ARPANSA’s Chief Executive Officer, framed the stakes in November 2025, when the transition to ANNPSR was formalised: “Robust regulatory frameworks are critical to the success of Australia’s nuclear-powered submarine programme,” she said. It’s worth reading that sentence closely. Hirth is not only talking about safety in the narrow sense. She is talking about the programme’s credibility as a whole. A regulatory failure – an enforcement gap, a mishandled incident, a loss of public trust – would damage the AUKUS submarine programme itself, at a moment when public and political scrutiny of that programme is already intense. In late May 2026, the government’s announcement that it would drop plans for a new-build Virginia class submarine in favour of a third second-hand vessel drew fresh criticism, sustained into June, that Australia was getting a poor deal from its AUKUS partners. It’s a reminder that the programme’s political tolerance for missteps, regulatory or otherwise, may not be large.
What other regulators can take from this
ANNPSR is an extreme version of a challenge that is becoming more common across regulatory practice: agencies being asked to oversee activities that do not yet fully exist at scale, in domains where there is no established regulated population and often no established public trust. Advanced nuclear technologies, autonomous systems, synthetic biology and parts of artificial intelligence governance all share elements of this pattern.
Three things ANNPSR is doing differently are worth other regulators’ attention. First, it is building capability ahead of the workload, rather than waiting for demand to force the issue – embedding staff with allied regulators years before the domestic caseload requires that expertise. Second, it has made consultation a legal precondition of licensing rather than a courtesy: under the Australian Naval Nuclear Power Safety Regulations 2025, the regulator must be satisfied that meaningful consultation with stakeholders, First Nations peoples, and the wider public has occurred before it can grant a licence. Third, and perhaps most unusually, ANNPSR has been candid in its own public documents about what it does not yet have: workforce, established processes, a settled framework for the activities still to come. That kind of institutional honesty is rare, and it is more likely to build legitimacy over time than a performance of certainty the regulator does not possess.
Back at the April education sessions, one unnamed attendee from the regulated community offered a reaction to the day that captured something ANNPSR appears to understand about its own position. “There are few opportunities that bring the regulated community together,” they said, “and acknowledging the role of the licensees in regulation is a positive step.”
It’s a modest observation that points to something larger. ANNPSR does not have decades of institutional trust to draw on. It cannot rely on a settled body of case law or an established professional culture shared with the people it regulates. What it has, instead, is the chance to build that relationship deliberately, from the first licence onward, treating the regulated community not as a problem to be managed but as a co-participant in the same objective. Whether that approach survives contact with a submarine programme moving at speed, under sustained political and public attention, will be one of the more instructive regulatory stories to watch over the next decade.
