The Australian government has signed a memorandum of understanding (MoU) with Anthropic, setting out plans to collaborate on artificial intelligence development, infrastructure, and safety under its National AI Plan. The agreement marks the first formal partnership under the government’s AI strategy and includes a focus on data centre infrastructure, energy supply, and research collaboration.
Under the MoU, Anthropic has committed to aligning any future Australian operations with the government’s recently released expectations for data centres and AI infrastructure developers, while also exploring investment in compute and energy capacity.
Infrastructure and energy in focus
A central component of the agreement is the role of infrastructure in supporting AI deployment. Anthropic signalled it is exploring investment in data centres and energy systems in Australia, including engagement with government on expanding electricity supply and transmission, particularly around firmed renewable energy.
The MoU also highlights the need to forecast infrastructure demand linked to frontier AI systems and to develop supply chains capable of supporting that growth. This aligns with broader government positioning of data centres as strategic assets underpinning AI adoption and digital economic growth.
Safety, research and workforce collaboration
The agreement places strong emphasis on AI safety and research collaboration, including cooperation with Australia’s AI Safety Institute and local academic institutions. Anthropic will share insights into emerging model capabilities and risks, while also contributing data to help the government track AI adoption and its economic impact across sectors such as natural resources, healthcare, and financial services.
The company also announced AUD 3 million in partnerships with Australian research institutions, including the Australian National University and the Garvan Institute, focused on applications such as genomics and disease diagnosis. Workforce development and skills were also highlighted, with the MoU outlining potential collaboration to support startups, small businesses, and public sector adoption of AI.
Government frames deal as signal to investors
Federal ministers positioned the agreement as part of a broader effort to attract AI-related investment while maintaining regulatory and societal safeguards. “This MoU sends a clear signal to Australians that we are open for business, where investment aligns with Australia’s priorities and Australian values,” said Minister for Industry and Innovation and Minister for Science Tim Ayres.
“The Australian Government and Anthropic are working together to harness AI responsibly – securing investment into Australia’s energy grid, driving economic resilience and safety for all Australians,” he said, adding that the agreement would support delivery of the National AI Plan and strengthen Australia’s position as an investment destination.
Assistant Minister for Science, Technology and the Digital Economy Andrew Charlton said the government was focused on ensuring AI delivers public benefit. “Australia needs technology partners who share our values and are committed to supporting our national interests, and this agreement is a positive step in that direction,” he said.
Non-binding agreement, but signals intent
The MoU is non-binding and does not commit either party to specific investments or procurement outcomes. However, it sets out a framework for future collaboration across infrastructure, research, and economic development.
Anthropic CEO Dario Amodei (above) met with Prime Minister Anthony Albanese in Canberra to formalise the agreement, as part of a broader push to deepen engagement in the Asia-Pacific region.
The company is also preparing to establish a local presence, including plans for a Sydney office.
Copyright emerges as unresolved issue
However, questions remain around copyright and content usage in AI development. According to the AFR, copyright was absent from the public announcements surrounding the agreement, despite being raised in discussions between Anthropic and Australian data centre operators including NEXTDC, AirTrunk and Canberra Data Centres.
AirTrunk founder and chief executive Robin Khuda described copyright as “the elephant in the room”, pointing to ongoing uncertainty over how AI companies will access and compensate for Australian content. The federal government is continuing consultations on potential reforms, while industry groups across publishing, music, and media have argued that existing frameworks can support commercial agreements.
Khuda suggested that subscription-based models could provide a pathway forward, ensuring content creators receive compensation as AI adoption expands.