91ɫ��Ƭ

Public law values must be protected as governments continue using technology in decision-making about people’s rights, obligations, entitlements and other interests, says Professor Janina Boughey, Director of the .

In the wake of the unlawful Robodebt scheme, and as governments have begun using AI-enabled tools, concerns are rising over the transparency and accountability of decisions, Prof. Boughey says.

In some cases, she warns, “governments themselves might not be able to explain how a system works”.  This is troubling, according to Prof. Boughey, because we are only able to hold governments to account for their actions if we understand what those actions are, and why decisions have been reached.

The Gilbert + Tobin Centre for Public Law brings together experts across law, technology and government to map where public law principles collide with automation and explore how safeguards can be built in.

The Centre’s researchers are currently particularly focussed on governments’ use of AI, how technology is being used to increase inequality, and government transparency in its uses of technology.

Increasing focus in technology in government decision making

Rapid digital adoption is driving an ongoing focus on automated decision-making (ADM) in government and the need for accountability, Prof. Boughey says.

“One would think that following the Robodebt Royal Commission Report, the government might be a bit cautious about using technology to make decisions with significant impacts on people’s lives. But that doesn’t seem to be the case.”

This focus has intensified since the Australian government started using ADM to elderly people receive.  

Similar plans are underway for the National Disability Insurance Scheme (NDIS). In May, the federal government introduced an that would authorise the use of ADM to determine individual eligibility for the NDIS. 

A into the Bill received more than 3500 submissions, which overwhelmingly raised concerns about the unintended consequences of using ADM to administer the NDIS. While the committee acknowledged these concerns in its , it nevertheless recommended that Parliament pass the bill, saying it would “secure a sustainable future” for the NDIS. 

Falling through the cracks

The shift toward digital services has delivered many clear benefits, Prof. Boughey says.

“There are some huge positives,” she says. “Technology can make the law and government services more accessible and much more efficient to deliver. 

“But government's use of technology also carries some significant risks.”  

Unequal access remains a key concern. “Certain sectors of the population who may not have easy access to technology, or might struggle to use it, might then struggle to access government services or complete these tasks.” 

When essential services move online without adequate support, the people who already face barriers risk falling further behind. What works well for many does not work equally for everyone. 

The highlights that digital exclusion still affects millions, especially people on lower incomes and those in regional communities.

Transparency and accountability

With new frameworks being developed in Australia and overseas, transparency, and by extension accountability, is at risk, Prof. Boughey says.  

The fact that governments themselves might not be able to explain how a system works creates a difficult problem, compounded by the technical complexity involved.

“Reading and interpreting legislation is quite a different skill set from understanding computer code,” Prof. Boughey says. “To check whether a system is working consistently with what the law says, you need to be able to do both. And I hazard a guess that there’s very few people who can both interpret legislation and understand computer code.”   

Robodebt’s lasting impact

That gap opens the door to systems that operate in ways no one fully understands, let alone supervises. Robodebt, an automated debt recovery program for welfare payments used by the federal government from 2016 to 2019, was the clearest and most devastating example of what can happen when automated decision systems drift away from legal requirements and human oversight when safeguards are insufficient. 

Ongoing reforms stemming from the show just how much work remains to rebuild trust in automated government systems.

Researchers at the Gilbert + Tobin Centre of Public Law are tackling these issues head on by exploring how safeguards can be built in, especially as AI makes the transparency and accountability challenges of automated decision making even more pronounced.  

Building safeguards

“We have brought together experts in law, technology and government to explore how these challenges can be dealt with and how to ensure public law values are protected in government decision making using technology,” Prof. Boughey says. 

This includes examining how to preserve fairness, how to protect transparency into automated processes, and appropriate avenues for review when decisions go wrong. 

“Technology is going to continue to develop, and governments are going to keep using it, and they should because of the benefits that it brings,” Prof. Boughey says. “Our goal is to make sure that it’s done in a way that’s fair, equal and just so that situations like Robodebt don’t happen again.”