Treasurer Jim Chalmers doubled down on a promise this week to conduct another review into artificial intelligence (AI) regulation, to determine whether the technology needs new laws or can fit into existing ones.
The government will accelerate a national development plan around how Australians can use the technology, Chalmers said at the government's productivity roundtable on Thursday.
AI could boost Australia's labour productivity by more than 4 per cent, but an expert warns the benefits may not be shared evenly, as Australia's peers around the world navigate the same challenges.
Australian Council of Trade Unions (ACTU) secretary Sally McManus said at the economic summit she was not calling for "over-regulating" AI, but argued for a worker-centric approach to rolling out the transformative technology.
Regulation of AI proved a major dividing line between employers and unions heading into the talks.
The ACTU called for the government to force employers to consult with staff before introducing new AI tools, while business groups warned that additional regulation could stifle adoption and reduce productivity gains.
In a major development, unions and the tech sector struck a deal to design a model that would see Australian creatives paid for their work when it is used to train AI.
McManus described the agreement with Tech Council chairman Scott Farquhar as a "breakthrough".
"There's agreement that we’re going to give this a real good go at coming up with a model that makes sure that people are actually paid for what they produce. So that’s a big thing," McManus said.
Business leaders also acknowledged that the risks of AI extend beyond job losses.
"I made the point that there’s risk not just for workers. Business owners are looking at their own future or extinction," said Australian Industry Group chief executive Innes Willox.
A Productivity Commission report released earlier this month estimated AI could add $116 billion to Australia’s economy over the next decade, but only if the government avoids stifling the technology with heavy-handed regulation.
The government has previously said it's working with industry to develop a National AI Capability Plan due by the end of the year, which will set out how to boost investment, grow industry capability, improve AI skills, including retraining workers.
Australia's cautious approach
The Productivity Commission has recommended the government hold off on introducing new AI-specific laws.
Instead, it says existing frameworks such as consumer protection, privacy, and anti-discrimination laws should be adapted first.
According to its Harnessing Data and Digital Technology interim report, the commission states AI-specific regulations should only be considered as a last resort.
The report also urges the government to pause its work on mandatory guardrails for high-risk AI until these gaps in existing law are fully understood.
It further stresses specific regulation should only be introduced if those frameworks cannot be adapted and if technology-neutral approaches are not workable.
The 'high stakes' of AI
According to the report, AI has the potential to lift labour productivity growth by 4.3 per cent over the next decade, but experts caution the benefits may be uneven, and warn that sidelining human oversight could undermine good governance.
Daniel Popovski, AI policy and advocacy lead at the Governance Institute of Australia, told SBS News the stakes are high with the rise of so-called 'agentic AI' — autonomous systems that act with little human oversight.
"Agentic AI is not just about producing content like chatbots do. These are autonomous systems run on very little human oversight. That raises new and urgent questions about accountability, transparency, and workplace governance that cannot be ignored," Popovski said.
Popovski said the conversation around AI is too often framed as one of job losses, overlooking its growth potential — provided there is a clear strategy in place.
"In reality, there is enormous potential for growth and augmentation if the transition is managed well. The challenge is making sure Australians are prepared and supported through these changes so that AI works for people, not against them," he said.
"We need a national AI strategy that sets out how the government will manage these challenges and opportunities.
"Good governance should be human-centric. It should support organisations that adopt AI responsibly and make sure the gains flow to all workers," Popovski said.
How the global landscape compares
Around the world, governments are taking different paths to manage the rise of AI. Some are prioritising protections for workers through up-skilling, while others are moving ahead with binding rules.
China is pushing for global cooperation, and the United Kingdom and New Zealand are aligning oversight with existing laws.
United States
The US AI action plan directs its labour department to fund rapid retraining for workers who lose jobs due to AI and to guide states on upskilling workers at risk of future displacement.
European Union
The EU is proceeding with the AI Act. Obligations for general-purpose AI models started on 2 August, with high-risk system requirements due in August 2026. A voluntary code of practice is already available to help companies prepare.
China
At the 2025 World AI Conference in Shanghai, China released a Global AI Governance Action Plan that calls for stronger international cooperation, inclusiveness and attention to environmental sustainability.
United Kingdom
The UK government has said it will consult the public on AI legislation and is aiming for a proportionate approach that works with existing laws, following its pro-innovation framework.
New Zealand
New Zealand has published its first national AI strategy. It aligns with OECD principles and encourages responsible use while aiming to accelerate private sector adoption and innovation.
