CANBERRA, 15 July 2026 — Prime Minister Anthony Albanese has unveiled the government’s National AI Plan, pairing major investment with new safety rules to manage the rapid rollout of artificial intelligence (AI) in Australia.
The centrepiece is the creation of the Australian AI Safety Institute, backed by $29.9 million. Launching in early 2026, the Institute will test high-risk AI systems, monitor emerging threats, and develop voluntary safety standards.
“We want to make sure all Australians benefit from AI,” Albanese said. “Our National AI Plan is all about capturing the economic opportunities of this transformative technology while protecting Australians from the risks.”
We're making sure AI works in Australia's interests. pic.twitter.com/mbGDfCijoB
— Anthony Albanese (@AlboMP) July 15, 2026
To support growth, the PM announced Microsoft’s A$25 billion investment in Australian AI and cloud infrastructure by 2029. The commitment includes expanding Azure AI supercomputing, strengthening cyber defences with the Australian Signals Directorate, and training three million Australians in AI skills. Microsoft will also collaborate directly with the new AI Safety Institute.
The government also released National Data Centre Expectations. Under the new rules, operators must underwrite new renewable energy, pay their full share of grid costs, use water sustainably, invest in Australian jobs and skills, and provide compute access to local start-ups.
Industry Minister Tim Ayres said the expectations will guide approvals and ensure Australia benefits from the projected $190 billion needed to triple computing capacity by 2030.
The plan also expands AI in government through the GovAI platform and a new AI Delivery team. It follows recent deals including OpenAI’s $7B Sydney data centre and an MoU with Anthropic.The government will work with states and industry to implement the new expectations, with the Safety Institute to release its first framework in early 2026.
