MRC Submission: Introducing mandatory guardrails for AI in high-risk settings

The MRC makes a Submission to the Australian Government's consultation process on ai regulation

Published: October 2024

The rapid growth in applications for AI has the potential to reverse the alarming decline in Australia’s productivity growth. A critical enabler (or inhibitor) of AI will be the regulatory environment.

 The Government appears to be moving to impose stringent restrictions on AI focussed on “preventing [imagined] harms before they arise” and requiring developers to address risks against “all foreseeable use cases by their clients”. This approach will crush innovation and is akin to banning the development of fast cars because drivers might speed, rather than making speeding illegal.

The MRC’s submission argues that the Government should step back from this unnecessarily burdensome approach. Australia should instead follow the approach of the previous UK Conservative Government, which would mean coordinating existing regulations and regulators to drive a technology-neutral approach that avoids stifling new innovations before they occur.