Anthropic signs an AI safety MOU with the Australian government, investing 3 million Australian dollars to support medical and education research

Anthropic CEO Dario Amodei traveled to Canberra, Australia, on March 31 to meet with Australian Prime Minister Anthony Albanese and officially sign a memorandum of understanding (MOU) for cooperation between the two sides, which will cover areas such as AI safety research, economic data sharing, and data center investment, to support the Australian government’s National AI Plan.

Sharing model capabilities and risk assessments with the AI Safety Institute

The core of the MOU is a cooperation framework with Australia’s AI Safety Institute: Anthropic will share capabilities of emerging models and risk research findings, participate in joint safety and safety assessments, and conduct research collaboration with Australian academic institutions. This arrangement is the same mechanism Anthropic previously established with safety research institutes in the United States, the United Kingdom, and Japan.

In addition, Anthropic will provide the Australian government with its “Anthropic Economic Index,” data used to track AI adoption across industries in Australia, its economic impact, and its influence on the workforce. Initially, it will focus on key sectors such as natural resources, agriculture, healthcare, and financial services.

A$3 million funds four research institutions, focusing on rare diseases and genetic diagnostics

Anthropic also announced that it will extend its AI for Science program to Australia by funding four Australian research institutions with A$3 million in Claude API credits:

Australian National University (ANU): using Claude to analyze gene sequencing data to assist in diagnosing rare diseases; ANU’s School of Computing will also incorporate Claude into its curriculum to cultivate the next generation of developers and scientists

Garvan Institute of Medical Research: two genomics research projects—one in collaboration with UNSW to translate human genetic variations into insights on disease mechanisms to discover new therapies; the other in partnership with the Population Genomics Center at Children’s Hospital, to automate the genetic analysis bottleneck in diagnosing rare genetic diseases in children

Murdoch Children’s Research Institute: applying to stem cell medicine programs to improve the identification of treatment targets for pediatric heart disease

Curtin University: Curtin’s Data Science Institute will leverage Claude to expand interdisciplinary research collaborations across fields such as health sciences, humanities, business, law, and engineering

Australian users’ ways of using Claude are the most diverse among English-speaking countries

Anthropic’s Economic Index data shows that among English-speaking countries, Australian users employ Claude for the most diverse range of task types, and tend to collaborate on high-skill tasks with complex prompts, covering management, business operations, and life sciences.

On the infrastructure front, Anthropic stated it is evaluating the possibility of investing in data centers and energy infrastructure in Australia, aligning with the Australian government’s recent expectations for data center development.

This MOU makes Australia the fourth government, after the United States, the United Kingdom, and Japan, to establish a formal AI safety cooperation framework with Anthropic.

This article about Anthropic and the Australian government signing an AI safety MOU and investing A$3 million to support medical and educational research first appeared on Chain News ABMedia.