– By Anna-Maria Arabia OAM
The Australian Academy of Science has sounded an alarm that policymakers can no longer afford to ignore. Its landmark Australian Science, Australia’s Future: Science 2035 report identifies 8 critical science domains where gaps in workforce, infrastructure and coordination will cripple our ability to meet future challenges: agricultural science, AI, biotechnology, climate science, data science, epidemiology, geoscience, and materials science.
The diagnosis is sobering. The prescription demands urgent attention from those who shape higher education policy.
A vision of science in 2035
Why does the Science 2035 report matter? Because capability gaps are not just a skills problem, they are a sovereignty problem.
Since 1990, 90% of global technological advancement has been rooted in fundamental science, yet Australia has spent decades allowing its STEM workforce to be determined haphazardly.
We are now at risk of compromising our ability to adequately respond to climate disruption, the critical minerals transition, or the AI revolution.
A nation that cannot generate its own scientific knowledge or secure diverse international research collaborators, becomes dependent on others to solve its problems, a dependency fraught with risk in a volatile world.
Future workforce needs
The report unambiguously shows that a critical misalignment persists: the current pipeline and study choices of students are not aligned with the needs of our future workforce. And declining STEM participation, shortages of in-field teachers, and unfocused international STEM mobility threaten national capability.
Fixing this requires active intervention: curriculum signalling; targeted incentives for students entering priority disciplines; alignment of workforce and immigration policy objectives; and partnerships between universities and industry to make career pathways visible and attractive well before students reach Year 12.
Universities are Australian assets
This is precisely where Australian universities must be understood as unique and strategic national assets, not credential factories. Universities can do something that no technical training program can replicate: they deliver research-led education.
When a student is taught by a practicing researcher working at the frontier of geoscience or epidemiology, they do not merely learn what is known, they learn how knowledge is created and challenged. That depth is exactly what the public and private sectors need when it confronts problems that have no established answers.
Industries that employ PhD graduates derive significant adaptive capacity from these employees through improved innovation, advanced problem-solving, and the ability to navigate complex, data-driven environments. And we know innovative industries are profitable and productive ones too.
The future with AI
A purely technical training response to the Science 2035 findings would be a mistake. Training people to operate today’s tools prepares them for yesterday’s challenges.
The AI landscape of 2035 will not be navigated by graduates taught only to use current software, it will be shaped by graduates who understand the underlying mathematics and can innovate when those tools evolve or become obsolete. Research-led science education builds that adaptability.
Follow the roadmap
The Science 2035 report gives policymakers something rare: a map of what needs to be done, backed by evidence. The response must match that ambition.
Sustained investment in university research and research-led teaching is not a cost, it is the essential infrastructure on which Australia’s future standard of living depends.
Anna-Maria Arabia OAM
Visiting Fellow, Australian National University
Former Chief Executive of the Australian Academy of Science (Oct 2016 to April 2026)
Image: Australian Academy of Science
Read more from Australian University Science: