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 291

Science & Technology Australia

Submitter group: Industry bodies.

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.

Argues that Australia must retain sovereign AI capability grounded in its research system, STEM workforce, and public compute infrastructure, and supports proportionate, technology-agnostic, risk-based AI regulation. It makes recommendations on research access to AI tools, decade-scale compute funding, STEM skills and data centre reporting.

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. Argues Australia should retain the capacity to understand, develop, adapt and govern AI rather than only adopting overseas technology.

    Australia will not capture these benefits simply by adopting AI technologies developed elsewhere. We must retain the capacity to understand, develop, adapt and govern these technologies ourselves.

    A program corrected obvious text-recognition errors in this quote for display.

  2. Argues sovereignty does not require reproducing the global AI supply chain but does require domestic expertise and infrastructure for informed choices.

    Australia cannot and should not seek to reproduce every element of the global AI supply chain. But we must have sufficient domestic expertise and infrastructure to make informed choices about the technologies we use

    A program corrected obvious text-recognition errors in this quote for display.

  3. Argues that private data centre and cloud capacity should not replace publicly funded, publicly accessible research compute.

    Australia’s rapidly expanding private data centre and cloud computing capacity presents an opportunity to strengthen our sovereign research infrastructure — but must not be seen as a replacement for publicly funded and publicly accessible compute capabilities.

  4. Says Australian research expertise and datasets in health, agriculture, climate, resources and other fields offer opportunities for domain-specific models.

    This is particularly relevant to the Committee's consideration of narrow and domain-specific models trained on Australian data. Australia possesses internationally significant research expertise and datasets in areas including health, agriculture, climate and environmental science, resources, astronomy, defence, biodiversity and many other fields.

  5. Recommends government, universities, schools and industry promote mathematics throughout education and address out-of-field teaching and declining higher-level participation.

    These skills start in school — with Australian students’ maths confidence and capability a particular concern. The Government, universities, schools and industry should work together to position mathematics as a vital subject throughout education and address the prevalence of out-of-field maths teaching, as well as a persistent decline in higher-level maths participation.

  6. Argues regulation and innovation should not be treated as competing objectives.

    Australia should avoid treating AI regulation and AI innovation as competing objectives. Trustworthy regulatory settings can be strong enablers through building greater certainty for researchers, businesses and the community.

Recommendations

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

  1. Recognise and fund sovereign AI capability.

    To secure Australia’s capability in world-leading research, expertise to enable effective regulatory approaches and development of bespoke AI applications in the national interest, the Australian Government should recognise the importance of securing Australia’s sovereign AI capability — and fund it accordingly.

    A program corrected obvious text-recognition errors in this quote for display.

  2. Take note of how AI is transforming the research process and ensure all Australian researchers have access to the appropriate AI tool they need for their research.

    The Australian Government should take note of how AI is transforming the research process and ensure that all Australian researchers have access to the appropriate AI tool they need to advance their research.

    A program corrected obvious text-recognition errors in this quote for display.

  3. Ensure researchers can access the compute infrastructure needed for internationally competitive or sovereign-critical work where there is a clear national benefit.

    Government investment and policy settings should work to ensure researchers can access the compute infrastructure required to undertake internationally competitive and / or sovereign-critical work where there is a clear national benefit.

  4. Deliver secure funding on a decadal scale, rather than short grant cycles, to national level compute capability.

    Deliver secure funding on a decadal scale, rather than short grant cycles, to national level compute capability.

  5. Assess current and future national compute requirements across research, government and industry, including public research infrastructure, university facilities, government capability and commercial providers.

    To ensure Australia’s research needs are met, the Government should undertake a coordinated assessment of current and future national compute requirements, including the respective roles of publicly funded research infrastructure, university-owned facilities, government’s own capability (such as those hosted by CSIRO, the Bureau of Meteorology and Defence) and commercial providers.

  6. Invest in the STEM skills pipeline from school to graduate level, including out-of-field maths teaching.

    the Australian Government should invest in bolstering Australia’s STEM skills pipeline. This should be done through targeted investments spanning school education through to graduate level — including support to address the urgent issue of out-of-field maths teaching.

  7. Make strategic investments in both AI specialists and AI-enabled professionals who use AI within their own fields.

    Australia should therefore distinguish between AI specialists, who develop and deeply understand these systems, and AI-enabled professionals, who require the skills to use them safely and productively within their own areas of expertise. The Committee’s recommendations should direct the Government to make strategic investments in both these areas.

  8. Adopt proportionate, technology-agnostic, risk-based AI regulation with strong human oversight and protections for privacy, security, transparency and accountability, while enabling responsible research, development and adoption.

    The Australian Government should adopt proportionate, technology-agnostic and risk-based AI regulation, with strong human oversight and protections around privacy, security, transparency and accountability, while enabling responsible research, development and adoption.

  9. Establish stronger national AI coordination and safety capability, with strong connections to the scientific and research community.

    As such, stronger national AI coordination and safety capability should be established, and these functions must maintain strong connections with Australia's broader scientific and research community.

  10. Require new AI infrastructure receiving Commonwealth facilitation to report energy sources, water use and demonstrated efficiency measures alongside its public-benefit case, consistent with data centre principles developed with states and territories.

    The Australian Government should require new AI infrastructure receiving Commonwealth facilitation to report energy sources, water use and demonstrated efficiency measures alongside its public benefit case, consistent with the national data centre principles being developed with the states and territories.

  11. Acknowledge the criticality of Australia's entire R&D and innovation system and, informed by the Ambitious Australia report, implement the policy and funding settings to sustain it.

    The Australian Government must acknowledge the criticality of Australia’s entire R&D and innovation system, which underpins our sovereign AI capability, but can also be supercharged by AI. Informed by the Ambitious Australia report, the Government must take steps to implement the policy and funding settings that will sustain Australia’s R&D and innovation system to drive future productivity, prosperity and sovereign resilience.

    A program corrected obvious text-recognition errors in this quote for display.

  12. Account for growing AI-related compute and data needs in government planning, policy and funding, notably the 2026 National Research Infrastructure Roadmap.

    High-performance computing and data capability already underpin research across disciplines and that demand will only continue to grow substantially as AI becomes embedded throughout research. Government planning, policy and funding settings — notably, the 2026 National Research Infrastructure Roadmap currently being developed by Department of Education — must take this into account.

    A program corrected obvious text-recognition errors in this quote for display.

  13. Protect research data and IP and enable movement between providers without financial or technical barriers.

    Australia’s research sector also requires strong protections for data and intellectual property and the ability to move data between providers without financial or technical barriers.

  14. Consider requiring hyperscalers, data centres and AI providers to train local AI and technical staff, support in-house research or contribute to a sovereign research fund, potentially through agricultural-style levies.

    There are several ways that this could be realised. Hyperscalers, data centres and AI providers could be required to contribute to training local AI and technical staff, to support in-house research teams, or to contribute to a sovereign research fund that would support Australia’s research capability — either targeted AI research or more broadly. This could operate in a way similar to current agricultural research levies.

    A program corrected obvious text-recognition errors in this quote for display.

  15. Consider low-cost academic and research access to provider services, coordinated through mechanisms such as NCMAS by research merit and national priorities, with appropriate technical support.

    Other possible mechanisms could be to provide low-cost access to services for academic and research institutions. Importantly, access should be coordinated through existing national mechanisms such as the National Computational Merit Allocation Scheme (NCMAS), ensuring compute is allocated according to research merit and national priorities, and is delivered alongside appropriate technical support.

  16. Have government, universities, schools and industry promote mathematics throughout education and address out-of-field maths teaching and declining higher-level participation.

    The Government, universities, schools and industry should work together to position mathematics as a vital subject throughout education and address the prevalence of out-of-field maths teaching, as well as a persistent decline in higher-level maths participation.

  17. Combine domestic skills investment with targeted skilled-migration pathways, at least in the short term.

    It is important to recognise that the required workforce will, at least in the short term, require a mix of domestic skills investment and targeted skilled-migration pathways.

  18. Give regulators access to deep technical expertise, including outside government, so regulation keeps pace with technology.

    Regulators need access to deep technical expertise, including expertise outside government, to ensure that regulatory frameworks keep pace with rapidly changing technology.

Read the original

In some quotes a program corrected obvious text-recognition errors for display.

Open in the interactive guide