The federal government is signalling a tougher approach to artificial intelligence safety, with Assistant Technology Minister Andrew Charlton indicating that new regulatory powers and stronger enforcement measures are under consideration. Prime Minister Anthony Albanese is expected to outline the government’s broader AI strategy next week, including its response to workforce disruption, copyright concerns and the significant water and electricity demands of data centres.
Charlton said the government is addressing AI safety across two areas. The first concerns immediate, everyday harms arising from the expanding use of AI, while the second focuses on frontier risks that could create entirely new categories of technological harm. He pointed to the Australian AI Safety Institute, which became fully operational earlier this year, as an important part of the government’s response.
According to Charlton, AI safety will be managed through a whole-of-government regulatory approach rather than by a single agency. “In the Albanese government, AI safety will be pursued through every relevant agency and regulator, backed by laws that already exist and are strengthened where they need to be by new powers and tougher enforcement.”
However, independent MP for Curtin Kate Chaney argues that the government’s response remains too slow and passive given the scale of AI’s potential impact. “The government is not moving at the speed we need for a technology with such far-reaching impacts. It's really taken a hands-off approach.”
The government’s National AI Plan, announced last year, rejected the introduction of a standalone AI Act. Instead, Labor chose to rely largely on existing legal and regulatory frameworks to address AI-related risks. Chaney does not necessarily support creating one comprehensive AI Act, but she believes the current policy response is inadequate and that new safeguards are urgently required.
Chaney has released an AI Discussion Paper containing 18 policy recommendations. Among her priorities is the establishment of a digital duty of care, which would place greater responsibility on technology companies and digital platforms to anticipate, prevent and mitigate harms caused by their products and services. The government has already pledged to legislate such a duty, but Chaney is pressing for faster action.
Her concerns are supported by Toby Walsh, Chief Scientist at the University of New South Wales AI Institute, who said existing laws would not be sufficient to deal with emerging AI risks. Walsh argued that the exact legislative structure is less important than ensuring that new laws are introduced. He said these protections could be contained in a single AI Act or spread across several pieces of legislation.
The debate highlights growing pressure on the Albanese government to move beyond speeches, plans and voluntary frameworks. While the government is signalling a willingness to strengthen enforcement, Chaney and AI experts argue that Australia needs concrete legislation and clear accountability measures now, before the technology’s risks become more difficult to control.