AI Inquiry Atlas

An unofficial, independent guide to the public records of the Joint Select Committee on Artificial Intelligence. Not connected with the Parliament of Australia or the committee. Its makers also made a submission (no. 481; see About). Summaries and scores are written by AI and can be wrong: check the original before relying on them.

Submission 266

Research Australia

Submitter group: Health.

An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.

Open in the interactive guide

Summary

Read the original submission for the full text.

Research Australia, the national peak body for health and medical research, supports AI adoption that advances health and sovereign capability while protecting safety, equity and trust. It recommends a national healthcare AI plan, unified health data strategy, beacon implementations, adaptive regulation, workforce investment and equity-focused design.

Key points

Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.

  1. Recommends a comprehensive national healthcare AI plan with a unified health data strategy and exemplar implementations.

    Research Australia recommends a comprehensive national healthcare AI plan, supported by a unified health data strategy, sustained investment in national research infrastructure, and exemplar implementations that translate responsible innovation into measurable health, social and economic outcomes

  2. Says Australia's health data is fragmented across systems, with barriers to discoverability, access and linkage.

    Australia’s health data remains fragmented across Commonwealth, state, territory and private systems, with barriers to discoverability, timely access and linkage.

  3. Says exporting Australian data while importing services can move value offshore and create dependencies.

    Exporting Australian data while importing services can transfer economic and strategic value offshore, weaken domestic capability and create supply-chain dependencies.

  4. Says rapid adoption of generative AI such as clinical scribes risks outpacing governance and regulation.

    The rapid, sometimes unregulated adoption of generative AI (including clinical scribes) risks outpacing provider governance and regulatory response.

  5. Says workforce capability and practical AI literacy are preconditions for safe adoption.

    Workforce capability is a precondition for safe adoption. Clinicians, researchers, managers, regulators and consumer partners need practical AI literacy as well as support to navigate ethical, legal, cultural, procurement and regulatory obligations.

  6. Says it considers social licence to be enabling infrastructure for health AI, alongside data, research infrastructure and workforce.

    Research Australia therefore considers social licence to be enabling infrastructure for AI, alongside data, research infrastructure and workforce capability.

Recommendations

Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.

  1. Develop a comprehensive, whole-of-sector national healthcare AI plan aligning research, regulation, health services, industry, jurisdictions and consumers, building on existing plans and roadmaps.

    Develop a comprehensive, whole-of-sector national healthcare AI plan that aligns research, regulation, health services, industry, jurisdictions and consumers, building on existing plans and roadmaps

  2. Establish a unified national AI and health data strategy with clear stewardship, secure access, interoperability, Australian data sovereignty and a public-facing approach to data and investment coordination.

    Establish a unified national AI and health data strategy, with clear stewardship, secure access, interoperability, Australian data sovereignty and a public-facing approach to data and investment coordination

  3. Create beacon implementations demonstrating safety, value, governance and scalability, including aged/primary care, childhood cancer, chronic/rare disease, clinical quality monitoring, regional, rural, remote and very remote health, and public health.

    Create beacon implementations in priority areas to demonstrate safety, value, governance and pathways to scale, including but not limited to aged care, primary care, childhood cancer, chronic disease, clinical quality monitoring, rare disease, regional, rural, remote and very remote (RRRvR) health and public health

  4. Strengthen proportionate, adaptive regulation and clinical governance, including transparent evaluation of safety, efficacy, bias, privacy, cybersecurity, commercialisation pathways and real-world performance.

    Strengthen proportionate, adaptive regulation and clinical governance, including transparent evaluation of safety, efficacy, bias, privacy, cybersecurity, commercialisation pathways and real-world performance

  5. Fund practical AI, data science, health economics and regulatory-navigation training for researchers, clinicians, services, custodians, public servants and regional, rural, remote and very remote workforces, including executive management and leadership.

    Invest in workforce capability, including practical AI, data science, health economics and regulatory-navigation training for researchers, clinicians, health services, data custodians, public servants and RRRvR workforces. This must include across the workforce, including executive management and leadership

  6. Build public trust and social licence through transparency, participatory governance, consumer partnership, consent reform, inclusive design and recognition of lived-experience expertise.

    Build public trust and social licence through transparency, participatory governance, consumer partnership, consent reform, inclusive design and meaningful recognition of lived-experience expertise

  7. Use public investment, procurement, infrastructure funding and commercialisation policy to retain Australian intellectual property, data value, talent and sovereign capability.

    Use public investment, procurement, infrastructure funding and commercialisation policy to retain Australian intellectual property, data value, talent and sovereign capability

  8. Ensure equity by design, addressing First Nations leadership, regional, rural, remote and very remote communities, women's health, people with disability, culturally diverse communities and other groups experiencing systemic barriers.

    Ensure AI adoption is equity by design, with explicit attention to First Nations’ leadership, regional, rural, remote and very remote communities, women’s health, people with disability, culturally diverse communities and other groups experiencing systemic barriers

  9. Co-design proposed health AI applications with consumers and evaluate them in real-world settings before broad deployment.

    These applications should be co-designed with consumers and evaluated in real-world settings before broad deployment.

  10. Have the national healthcare AI plan define priorities, responsibilities, governance, investment settings and success measures, aligning Commonwealth agencies, jurisdictions, health services, researchers, industry and communities.

    The plan should define priorities, responsibilities, governance, investment settings and measures of success, while aligning Commonwealth agencies, states and territories, health services, researchers, industry and communities.

  11. Connect the plan with the National Health and Medical Research Strategy, Ambitious Australia, the Productivity Commission's data/digital agenda and national research infrastructure planning, avoiding an isolated initiative.

    It should connect with the National Health and Medical Research Strategy, Ambitious Australia the Productivity Commission’s data and digital technology agenda and national research infrastructure planning rather than create another isolated initiative.

  12. Include clear implementation milestones, lead agencies, shared accountability across jurisdictions and independent monitoring and evaluation in the plan.

    The plan should include clear implementation milestones, lead agencies, shared accountability across jurisdictions and mechanisms for independent monitoring and evaluation

  13. Require beacon exemplars to demonstrate best practice, knowledge transfer and adoption confidence, with clear public value, mature governance, measurable outcomes and pathways to scale across metropolitan, regional, rural and remote settings.

    Beacon implementations should demonstrate best practice, enable knowledge transfer and build confidence in adoption. Priority exemplars should have clear public value, mature governance, measurable outcomes and pathways to scale across metropolitan, regional, rural and remote settings.

  14. Draw on collaborative models such as genomics-sector alliances integrating research and clinical practice.

    Australia should also draw on collaborative models such as the genomics sector’s alliance-based approach to integrating research and clinical practice.

  15. Include person-centred governance, nationally agreed standards, secure access and clear stewardship in the data strategy, recognising Indigenous Data Sovereignty, consumer control and visibility for under-counted populations.

    A unified national AI data strategy should establish person-centred governance, nationally agreed standards, secure access mechanisms and clear stewardship while recognising Indigenous Data Sovereignty, consumer control and the need to make under-counted populations visible in health systems.

  16. Develop a National Health and Medical Data Infrastructure Strategy mapping assets, harmonising governance, building common models and interoperability, and enabling secure, ethical linkage across clinical, administrative, genomic, phenomics, imaging and research datasets.

    Recent Research Australia positions also support a National Health and Medical Data Infrastructure Strategy that maps existing assets, harmonises governance, builds common data models, improves interoperability and enables secure, ethical linkage across clinical, administrative, genomic, phenomics, imaging and research datasets.

  17. Invest in trusted research environments, secure sovereign compute, synthetic-data governance, linkage, RRRvR datasets, registries, regional access and platforms such as Health Data Australia; reduce burdens for smaller teams, services, RRRvR communities and startups.

    This should include investment in trusted research environments, sovereign and secure compute, synthetic data governance, data linkage capability, RRRvR-relevant datasets, national registries, regional data access models and platforms such as Health Data Australia, while reducing the burden and uncertainty that currently prevents smaller research teams, health services, RRRvR communities and startups from using data for public benefit

  18. Support secure Australian infrastructure, narrow domain-specific models trained on Australian data, and commercialisation pathways that retain intellectual property, capability and value in Australia.

    Government should support secure Australian infrastructure, narrow domain-specific models trained on Australian data, and commercialisation pathways that retain intellectual property, capability and value in Australia.

  19. Support nationally coherent, proportionate, adaptive regulation with practical provider guidance, supply-chain accountability, post-market monitoring and independent evaluation in real-world clinical settings.

    Australia needs proportionate, adaptive and nationally coherent regulation supported by practical guidance for health providers, clear accountability across the AI supply chain, post- market monitoring and independent evaluation in real-world clinical settings.

  20. Treat privacy, cybersecurity and consent as core AI infrastructure.

    Privacy, cybersecurity and consent must be treated as core AI infrastructure.

  21. Consider consent, privacy law, ethics approval, commercialisation pathways, software-as-a-medical-device requirements and the emerging AI safety standards framework early.

    Research Australia supports early consideration of consent, privacy law, ethics approval, commercialisation pathways, software-as-a-medical-device requirements, and the emerging AI safety standards framework.

  22. Simplify participant information and consent, guide ethics committees more clearly, manage commercial data uses transparently and strengthen trusted, secure analysis environments.

    Research Australia supports simplified participant information and consent processes, clearer guidance for ethics committees, transparent management of commercial uses of data, and stronger support for trusted and secure analysis environments

  23. Anticipate demand distortion from imported health beliefs and model outputs that do not reflect Australian evidence, policy or population needs.

    Regulatory settings must also anticipate demand distortion from imported health beliefs and model outputs that do not reflect Australian evidence, policy or population needs.

  24. Ensure health AI deployed in Australia is demonstrably suitable for the Australian context, transparent about material limitations and subject to appropriate clinical oversight.

    Health AI deployed in Australia should therefore be demonstrably suitable for the Australian context, transparent about material limitations and subject to appropriate clinical oversight.

  25. Consider an adoption-support model similar to AI Adoption Centres, tailored to the health and medical research ecosystem and accessible to organisations with limited internal capability.

    Government should consider an adoption- support model similar to AI Adoption Centres, tailored to the health and medical research ecosystem and accessible to organizations with limited internal capability.

  26. Invest in clinician-researchers, data scientists, implementation specialists, health economists and regulatory experts, alongside AI literacy across the broader health workforce.

    Realising the benefits of AI will require investment in clinician-researchers, data scientists, implementation specialists, health economists and regulatory experts, as well as AI literacy across the broader health workforce.

  27. Involve consumers, patients, clinicians, researchers and diverse communities in AI development, evaluation, implementation and governance through co-design, participatory mechanisms and transparent communication.

    Consumers, patients, clinicians, researchers and diverse communities should be meaningfully involved in the development, evaluation, implementation and governance of AI through co- design, participatory mechanisms and transparent communication.

  28. Give particular attention to privacy, secondary data uses, representative datasets, algorithmic bias, explainability, accountability, cybersecurity and accessibility; complement trusted research, clinical and community relationships.

    Particular attention is required to privacy and secondary uses of health data, representative datasets, algorithmic bias, explainability, accountability, cybersecurity and accessibility. AI should complement, rather than undermine, trusted relationships between researchers, health professionals, patients and communities.

  29. Include public engagement, evaluation and trustworthy governance investment in national AI capability rather than treating it primarily as regulatory compliance.

    Investment in public engagement, evaluation and trustworthy governance should form part of Australia's national AI capability rather than being treated primarily as regulatory compliance.

  30. Embed Indigenous Data Sovereignty and Indigenous Data Governance in the development and use of AI, in line with Closing the Gap Priority Reform 4.

    In line with the Closing the Gap National Agreement, Priority Reform 4, Indigenous Data Sovereignty and Indigenous Data Governance should be embedded in the development and use of AI.

  31. Give First Nations peoples meaningful authority over their data and knowledge and related AI systems' design and accountability, recognising collective interests and community-defined benefits.

    First Nations peoples should therefore be more than data subjects or end-users. They should have meaningful authority over how data concerning their peoples, communities and knowledges are created, governed, interpreted and used, and how AI systems using those data are designed 18 Prehn (Worimi), J., Carlson (Aboriginal), B., Walter (Palawa), M., Lovett (Wongaibon/Ngiyampaa), R., & Maher (Yamatji, Noongar, Kija), B. (2026). Indigenous Data Sovereignty and the Governance of Artificial Intelligence: Toward Equitable Benefit for Indigenous Peoples. Journal of Sociology, 0(0). Research Australia Page 8 Response to Artificial Intelligence Consultation and held accountable. This requires recognition of collective as well as individual interests and community-defined understandings of benefit.

  32. Pursue equity through culturally safe governance, consumer partnership, digital inclusion, appropriate reimbursement and local relevance; consider who defines measures, representation, outcomes and benefits.

    More broadly, Australia should pursue equity by design. Digital and AI tools can widen existing disparities if developed without culturally safe governance, consumer partnership, digital inclusion, appropriate reimbursement pathways and local relevance. Bias cannot be addressed solely as a technical problem of improving datasets; governance must also consider who determines what is measured, how communities are represented, what outcomes are valued and who ultimately benefits.

  33. Make RRRvR AI approaches place-based and locally led wherever possible, supported by infrastructure and workforce investment, evaluating community-defined outcomes alongside conventional clinical and economic measures.

    AI approaches for RRRvR communities should therefore be place-based and locally led wherever possible, supported by appropriate infrastructure and workforce investment, and evaluated against community-defined outcomes alongside conventional clinical and economic measures.

  34. Provide long-term investment in health research and innovation enablers: workforce, data, infrastructure, discovery and implementation science, health services research, health economics, regional research and commercialisation capability.

    It requires long-term investment in the enablers of health and medical research and innovation: workforce, data, infrastructure, discovery science, implementation science, health services research, health economics, regional research capacity and commercialisation capability.

  35. Lift national R&D investment measurably; strengthen NHMRC and MRFF funding, release available MRFF funds, fund full research costs and bridge discovery, clinical development, implementation and market readiness.

    Research Australia supports a measurable pathway to lift national R&D investment, stronger funding for NHMRC and MRFF programs, release of available MRFF funding, funding for the full cost of research, and funding models that bridge the "valley of death" between discovery, clinical development, implementation and market readiness

  36. Use government procurement and public investment to support Australian-developed health technologies, sovereign manufacturing, secure data infrastructure and local commercialisation.

    Government procurement and public investment should be used to support Australian-developed health technologies, sovereign manufacturing, secure data infrastructure and local commercialisation pathways.

  37. Invest in trusted data environments, interoperable health data systems, sovereign compute, Australian-trained models, evaluation and regulatory pathways, and translation of AI innovations into routine clinical practice.

    For AI specifically, this includes investment in trusted data environments, interoperable health data systems, sovereign compute capability, Australian- trained models, evaluation and regulatory pathways, and mechanisms that support the translation of AI innovations from research into routine clinical practice.

  38. Consider an Australian BARDA equivalent potentially aligned with the Australian CDC, stronger IP support, industry PhDs and placements, industry/philanthropy co-investment, and end-to-end reimbursement, health technology assessment and adoption support.

    This includes considering models such as an Australian equivalent to Biomedical Advanced Research and Development Authority (BARDA) which could be aligned to 19 Draft National Health and Medical Research Strategy 2026–2036, Enablers and Enabling Initiatives (Workforce; Funding; Data and Advanced Technology; Infrastructure). Research Australia Page 9 Response to Artificial Intelligence Consultation the role of the Australian Centre for Disease Control, stronger IP support, industry PhDs and placements, co-investment models with industry and philanthropy, and end-to-end support for reimbursement, health technology assessment and adoption in health services.

Read the original

Open in the interactive guide