Australia will follow in the footsteps of the European Union, US, UK, Japan, Canada and other major economies by establishing its own AI safety institute, the federal government announced on Tuesday.
The new agency, which will be established in 2026 under the Department of Industry, Science and Resources, will monitor and address potential risks from AI as it becomes more prevalent in economic and government activities.
The institute will also ensure that AI companies comply with Australian law, support regulatory improvements across government and “advise where laws need to be updated,” Labor said.
Industry, Innovation and Science Minister Tim Ayers said while AI could “boost industry, increase productivity and raise living standards for all Australians”, the Commonwealth needed to protect Australians “from any misuse”.
“The institute will become the government’s center of AI safety expertise,” he said.
“It will be operated with transparency, responsiveness and technical rigor so Australians can use this innovative technology with confidence and safely.”
Dr Andrew Charlton, under-secretary for science, technology and the digital economy, said the government was “working closely with industry, business, trade unions and civil society to accelerate the uptake of AI in a safe and responsible way”.
The new institute will also collaborate with Australia’s National AI Center and other countries’ AI safety institutes, the government said.
‘Great move’: Supporters welcome AI announcement
The Australian AI Safety Institute (AISI) announcement follows long-standing support from numerous industry experts, associations and some politicians.
Industry bodies including the Business Council of Australia (BCA) and the Australian Academy of Technology, Science and Engineering (ATSE) had previously called on the government to establish AISI.
Greg Sadler, spokesperson for Australians for AI Safety, welcomed the government’s decision and said the nation’s prosperity and security depended on how it navigated advances in AI technology.
“The world-leading AI Safety Institute will provide Australia with the technical expertise to understand advanced AI, help prevent serious risks and put us on the path to global leadership,” he said.
Toby Walsh, chief scientist at the University of New South Wales AI Research Institute and member of the ACS AI Ethics Committee, who previously said Australia “desperately” needs AISI, reacted to Tuesday’s news with a simple “Bravo!”
Minister for Industry, Innovation and Science Tim Ayers (centre) says AISI will help Australians use AI safely. Image: Tim Ayers/LinkedIn
Other AI researchers also welcomed the move, but are awaiting further details on how the Australian lab will operate.
Associate Professor Michael Noetel from the University of Queensland said the establishment of the institute was a “great move” but experts would “closely monitor the details”.
“One of the most important things to get right is leadership,” he said.
“Other parts of government are focused on current AI risks and accelerating AI adoption.
“AISI must be led by someone with credibility in Silicon Valley who values the security of frontier AI development.
“They will need to attract and retain talent and negotiate with AI companies to gain access and transparency.”
Dr Alexander Saeli, an AI governance and risk researcher at the University of Queensland and the Massachusetts Institute of Technology, said the institute would need funding not only to attract talent but also to “access the computing resources needed to explore cutting-edge AI systems”.
“You also need a clear mission statement that sets the organization’s focus on frontier AI risks,” he said.
Opposition warns of ‘mixed messages’ from government
Shadow Minister for Industry and Innovation Alex Hawke said the Coalition government wanted Australians to be protected from AI risks and “supported to access the benefits of AI”, but insisted the Australian AISI announcement “does not make up for Labor’s hesitation”.
“AI requires serious and coordinated national leadership,” he said.
“Instead, we’re getting mixed messages from the government, including from Dumping Industries Minister Ed Husik,” he added, noting that former ministers continue to publicly support AI-only legislation.
The previous Labor government proposed mandatory guardrails for high-risk uses of AI under Husich, but the re-elected Albanon government was warned to suspend work on such guardrails and reportedly abandoned plans for AI-specific legislation such as the AI Act.
“AI must not become a Trojan horse for the trade union movement to provoke further industrial interference,” Hawke added, referring to Minister Ayers, himself a former trade union official, calling on trade unions to participate in discussions about the use of AI in the economy.
The Australian Council of Trade Unions (ACTU), which has been campaigning for workers to share in the benefits of AI after some workers have had their “content stolen, work disrupted and rights violated”, said it welcomed the government’s AISI announcement.
ACTU Assistant Secretary Joseph Mitchell said the establishment of AISI was an “important first step” in ensuring AI models, including those developed overseas, comply with Australian law and standards.
“The rapid development of these models has taken away too many livelihoods,” he says.
“The first step to sharing benefits is to protect yourself from potential harm.”
The government added that protecting Australians from the potential harms of AI is also a “key pillar” of the next national AI plan, expected to be released by the end of 2025.
Labour’s AI Plan for the Australian Public Service, released earlier this month, revealed plans to appoint a chief AI officer for every federal government agency in 2026.
