Australia urged to create AI sovereign wealth fund for citizens

Context Australia is courting major AI firms such as Microsoft and Anthropic, which seek stable locations for massive data centres powered by renewable energy. These firms need space, clean power and a predictable regulatory environment to train the large models that underpin services like chatbots and image generators. Public scepticism has grown as communities worry about the environmental toll of data centres and the fairness of profit distribution.

Key Facts Norway’s sovereign wealth fund, built from North Sea oil revenues, is now valued at over A$3 trillion and serves as a model for converting natural resource wealth into public benefit. Courts in several jurisdictions have ruled that the mass collection of web content to train AI models constitutes illegal scraping, a criminal enterprise that undermines creators’ rights. An AI sovereign wealth fund would require ongoing profit shares from any company that trains its AI models in Australia, capturing value wherever those models are later deployed.

What It Means By linking profit sharing to the location of model training, the fund would give Australians a direct stake in the AI boom while incentivising firms to respect local laws and data rights. Revenues could support creators whose work is used without consent, fund retraining for jobs displaced by automation, and invest in renewable infrastructure that powers the data centres. Policymakers would need to define the profit‑share rate, set up transparent governance, and ensure compliance with existing tax and copyright frameworks. Other nations are already experimenting with similar ideas; South Korea is debating an AI profits fund to support startups and social safety nets, while US think‑tanks have floated a “token tax” on AI usage. The next step to watch is whether the federal government releases a formal feasibility study or pilot scheme for the AI wealth fund within the next 12 months.