The Federal Government has received warnings over its use of AI decision-making. 

A Senate committee has expressed concern over the replacement of human discretion with automated systems, following the introduction of new regulations by Home Affairs Minister Clare O’Neil and Agriculture Minister Murray Watt. 

These regulations allow certain decisions, traditionally made by ministers and their officials, to be automated - a move that has not gone unnoticed.

A recent report by the Committee for the Scrutiny of Delegated Legislation notes the potential for serious issues.

“The committee considers that the use of an automated decision-making process may operate as a fetter on discretionary power by inflexibly applying predetermined criteria to decisions that should be made on the merits of each individual case,” it warned. 

The regulations introduced by O’Neil pertain to national security restrictions on visa holders seeking to study certain technologies. 

Where previously the minister had discretion over exemption applications, now an AI program will make these determinations. 

Similarly, Watt's regulation involves AI in deciding when individuals involved in biosecurity concerns must provide documents or information.

Both ministers argue the changes are needed to boost efficiency. However, the Senate committee has sought further clarification on how these AI systems will function, including the safeguards in place to ensure the continued exercise of ministerial discretion and the integrity of the merits-review process.

Senator Paul Scarr, acting chair of the committee, has voiced ongoing concerns about the balance between technological advancement and the preservation of fundamental legal safeguards.

“We’re having to repeatedly raise the same scrutiny concerns… the more substantial the decision, the greater impact it has on people’s rights and liberties and the more concerned we are,” he told reporters.