The Albanese government’s ambition to develop ‘sovereign AI’ capability in Australia has already piqued the interest of local and multinational technology firms hoping for a piece of the action, despite uncertainty and confusion over what the term actually means.
In a foreword to the National AI plan released this week Industry and Innovation Minister Tim Ayres said “building sovereign capability in AI” will help drive Australia’s economic resilience and competitiveness. But which companies and technologies count as sovereign was not explicitly defined.
What counts as sovereign capability is also different across government. The Defence Industry Development Strategy, which was released in February 2024, argues that “only in limited circumstances is Australian ownership critical to sovereignty”.
The National AI Plan also flags the ongoing roll out of “sovereign AI for the public service” through GovAI, which has access to onshore instances of OpenAI’s ChatGPT and Anthropic’s Claude.