google.com, pub-8701563775261122, DIRECT, f08c47fec0942fa0
Australia

‘Better world’: inside Australia’s pact with AI giant

1 April 2026 16:07 | News

A US tech giant has agreed to cooperate with Australia on AI safety as its CEO stresses the importance of regulating the controversial technology.

Anthropic CEO Dario Amodei arrived in Canberra to sign a memorandum of understanding at a meeting with Prime Minister Anthony Albanese on Wednesday as part of the government’s bid to encourage the responsible development of artificial intelligence.

Under the agreement, the company will share findings about the risks and capabilities of AI, collaborate with research institutions and engage in safety and security assessments as part of its commitment to work with the Australian AI Security Institute.

Media were allowed to watch the fireside chat, but news photographers were rarely stopped at the door due to restrictions on visual coverage. Anthropic declined to comment on the news photography ban.

Anthropic is the maker of the cutting-edge chatbot Claude.

The company has agreed to support the local AI ecosystem, collaborate on the development of the technology and ensure its future Australian operations align with the government’s expectations for data centers and AI infrastructure developers.

Compared to Google or Chat GPT parent Open AI, Anthropic has been more outspoken about the need for guardrails around the technology.

At Parliament House, Mr Amodei warned of the consequences if AI falls into the hands of countries with advanced surveillance systems.

“They could really go in the direction of a panopticon here,” he said, referring to a philosophical concept often applied to surveillance situations.

“On the international stage, I see this as a military competition.

“AI is a powerful technology, and I don’t want autocracies to be militarily stronger than democracies.”

Dario Amodei believes that risks must be reduced to reap the benefits of artificial intelligence. (PR IMAGE PHOTO)

Although the CEO referenced China in his comments, the company’s agreement with Australia came after Anthropic filed a lawsuit against the US Department of Defense.

This was triggered by the company’s insistence on implementing security measures to prevent the military from potentially using its technology for mass domestic surveillance and fully autonomous weapons.

The Pentagon has since identified Anthropic as a supply chain risk and banned U.S. government contractors from using the company’s technology for the military.

“I am the last person to deny that there are all these risks,” Mr. Amodei said.

“But the benefits are real, and if we can reduce the risk we could have a much better world.”

The Australian government announced a deal with Anthropic as a way to capture AI opportunities while keeping its citizens safe.

Stock image of artificial intelligence applications
Australia needs to make sure all chatbots are secure, not just one, an AI expert says. (Jono Searle/AAP PHOTOS)

But University of NSW computer science professor Toby Walsh said Australia needed to bring in other companies if it wanted its plans to work.

“Just ensuring Anthropic models are safe will not be good enough,” he told AAP.

“You want to make sure that Google’s and OpenAI’s and everyone else’s models are secure, too.”

He called on Australia to support AI investment with real funding rather than relying on a non-legally binding memorandum of understanding.

The digital industry’s peak body has welcomed the Anthropic agreement as a significant step in developing AI talent under Australia’s national AI plan.

“Global partnerships are critical, but they need to be based on national priorities to ensure the benefits of AI are realized across the Australian economy and society,” said Elizabeth Whitelock, chief executive of the Australian Information Industry Association.

In June 2025, the federal government announced a $20 billion deal with Amazon Web Services to fund data center infrastructure, while its NSW counterpart on Friday backed a plan for 15 data center projects worth $51 billion.

Questions have been raised about its sustainability due to concerns about the large amount of resources the data center industry uses.

artificial intelligence agreement
The data center industry is under scrutiny due to concerns about the amount of resources it uses. (Steven Markham/AAP PHOTOS)

According to the Australian Water Services Association, some large data centers use up to 40 million liters of water per day; This amount is enough for 80,000 Australian homes.

The Climate Council says they also require a significant amount of land, produce a lot of heat and can create noise pollution that affects nearby communities and animals.

“We want to make sure that this huge interest in data center investment in our country comes with obligations, including natural resource management, of course water, energy and a number of related factors,” Finance Minister Jim Chalmers told reporters.


AAP News

Australia’s Associated Press is the beating heart of Australian news. AAP is Australia’s only independent national news channel and has been providing accurate, reliable and fast-paced news content to the media industry, government and corporate sector for 85 years. We inform Australia.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button