Research Australia Response to the Consultation by the Department of Industry, Science and Resources on Safe and Responsible AI in Australia

This submission responds to the Department of Industry, Science and Resources’ June 2023 discussion paper Safe and Responsible AI in Australia, answering selected consultation questions with a focus on artificial intelligence (AI) in therapeutic goods and in health care more broadly. Research Australia maintains that the Therapeutic Goods Administration (TGA), which already regulates software as a medical device using a risk-based framework and leads international regulatory cooperation, is the most appropriate body to govern AI in therapeutic goods, and that AI is often best regulated as part of the product rather than separately. It proposes a two-tier model: whole-of-government principles based on Australia’s eight voluntary AI Ethics Principles, with detailed implementation by the regulator closest to each industry. It recommends that the Department of Health and Aged Care, which is already reviewing electronic clinical decision support in general practice, assess current and potential uses of AI across the health and aged care systems, and that the Government fund research on AI risks and opportunities in health, aged care, transport and education. It supports a risk-based approach, using the ConsultNote.ai tool as a case study to distinguish lower-risk uses such as drafting referral letters from higher-risk treatment advice, and recommends nationally consistent transparency across public, private and not-for-profit health services that makes clear when AI supports, rather than replaces, a clinician.