There has been a recent flurry of AI-related policy released, most notably the National AI Plan released on December 2 by the Department of Industry, Science and Resources (DISR). The plan takes a whole-of-government approach and covers many challenges Australia faces as it seeks to catch up with other nations in the global AI race. It contains few specific resource commitments, actions or timelines, and thus is in some sense an “aspirational” document.
One concrete yet high-level announcement was the creation of a National AI Safety Institute. With a budget of $30 million, the AI Safety Institute is tasked with monitoring the development of AI and advising industry, agencies and ministers.
For healthcare, the DISR Plan steps back from legislating the 10 “mandatory guardrails” that were on the table last year. Instead, it recognises that there are already existing regulations with governance processes in the different industry sectors, that should be the source of AI guardrails in the first instance. The AI plan does reserve the right to craft national regulation or legislation for AI safety where needed.
Healthcare in this sense is well-placed, with regulators and agencies including the TGA, Ahpra, and the Commission for Safety and Quality all focussing on AI, developing in-house expertise and taking a stake in AI governance. It could be argued that healthcare is ahead of the pack in AI governance compared to most other sectors, although much remains to be done, and the nature of the task continues to evolve. From that perspective, it will be essential for the new AI Safety Institute to work closely with different sectors on their specific needs, and to especially focus on healthcare, which is tackling issues of safety, quality and ethics early and head on.
Others have been disappointed in the Plan, noting it does not substantively address the issue of ensuring we have a national and sovereign AI capability, and that local industry is supported to become internationally competitive.
The AAAiH is in the process of developing its national AI Implementation Roadmap, scheduled for release in the first quarter of 2026. Part of that process will involve a full analysis of the National AI Plan, as well as healthcare-specific policy announcements. This will ensure that we provide timely and informed sector guidance on the next steps to the appropriate adoption of safe, secure, effective and ethical AI across healthcare. The Implementation Roadmap project is jointly funded by Macquarie University and the Digital Health Cooperative Research Centre.