In a context where the AI technologies used and experienced by Australians are predominantly designed and developed elsewhere, the Academy believes that the successful development of Australian AI capabilities is a vitally important policy objective. The National AI Plan is therefore a significant step in framing national priorities and directions.
In considering Australia’s sovereign AI capabilities, the Academy calls on the Government to Back Australian Thinking. This requires recognising that national capability in AI is not limited to technical infrastructure and computational capacity, but also depends on Australia’s intellectual, cultural and social capability to design, deploy and govern AI systems in the public interest. Just as economic and industrial sovereignty depends on domestic capability in manufacturing, energy and critical technologies, Australia’s AI sovereignty depends on sustained investment in the research, skills and knowledge systems that enable AI to be developed, adapted and governed in Australian contexts, for Australian communities.
Inclusion, First Nations priorities & the public interest
The Academy notes and supports the plan’s emphasis on inclusion, including the significance of AI for the challenge of the Closing the Gap targets for First Nations digital inclusion. This is an area where significant progress is now being made but where a much more sustained effort is required to ensure AI supports, rather than undermines, Indigenous data sovereignty and self-determination. Humanities research plays a critical role in relation to Indigenous knowledges, languages, governance frameworks and community-led technology design.
For Australia to maximise the potential benefits of AI and minimise its harms, it is essential that its development is responsible and inclusive. AI systems that are deployed in education, health, employment, justice and public administration will only be effective and trusted if they are grounded in culturally informed understandings of Australian communities and institutions. Achieving these objectives requires deeper engagement with the humanities and social science expertise across policy design, research funding, industry partnerships and workforce development.
Stronger linkages with university research capabilities are needed
Australian universities have developed substantial cross-disciplinary capabilities in the development, application, evaluation, and regulation of AI technologies. Australia also has world-leading expertise in the specific contexts where AI’s purported benefits and harms arise, from education and work to health, media, law and democratic institutions.
However, these strengths could be more closely and systematically linked to AI policy and industry. Stronger, more durable linkages between government, industry and university research will be essential if the National AI Plan is to deliver sovereign, inclusive and responsible AI capability. Engagement with humanities disciplines in particular will be critical: issues of inclusion, legitimacy, accountability and trust cannot be addressed solely through technical safeguards, important though these are.
Australia has internationally recognised strengths in humanities fields such as communication and media, languages, creative arts, philosophy, and history. These disciplines are essential to developing innovative AI systems, understanding and addressing their limitations, and evaluating their impacts in real-world contexts. Humanities practitioners also work closely with communities and end-user organisations all over Australia and are well-placed to support AI development that responds to their needs and concerns.
Humanities expertise is integral to future AI technologies
There is a growing international recognition of the importance of humanities disciplines for the future development of Generative and Agentic AI systems. As AI systems are increasingly expected to communicate, reason, translate and make decisions in complex and diverse social environments, their effectiveness will depend on how well they are suited to particular cultural, linguistic and social contexts.
Cultural and social dimensions must be embedded at all layers of AI systems and throughout the AI lifecycle: in the formulation of problems, the design of systems and data selection, their translation into applications, and their evaluation and deployment. Achieving this will require a skilled Australian workforce able to connect humanities and social science expertise with a strong understanding of the technological dimensions of AI systems.
This workforce challenge also highlights the importance of Jobs and Skills Australia’s forthcoming National Skills Taxonomy (NST) which is not referenced in the National AI plan. As the Academy has previously argued in its submission to Jobs and Skills Australia, skills are not merely discrete or technical competencies, but purpose-driven human capabilities developed through learning, practice and experience, and exercised in specific social and organisational contexts. For AI-enabled roles in particular, capabilities such as critical reasoning, ethical judgement, communication, cultural literacy and systems thinking must be understood as core skills rather than supplementary attributes. A national skills taxonomy that adequately recognises these capabilities—alongside technical expertise—will be essential to enabling employers, educators and policymakers to identify, develop and deploy the workforce required for responsible and effective AI development in Australia.
Successful national AI development will depend on cross-sector collaboration & meaningful public debate
Harnessing AI to achieve national economic, environmental, social, and cultural objectives represents a complex policy challenge. Experience to date shows us that AI systems have many positive benefits as well as many challenges. Issues such as responsibility, accountability, safety, and inclusion must be understood as cultural and social issues, as well as technical and business challenges.
Successful national AI development will demand co-ordinated collaboration across disciplines and sectors, including industry, business, community and consumer organisations, governments and regulators at all levels.
Public debate about AI in Australia is often polarised, reflecting legitimate concerns, alongside optimism about productivity and innovation. There are most certainly gaps in understanding among citizens, users, industry and government decision makers, at any point in time, and both the technologies and debates continue to rapidly evolve. Responsible, inclusive and ethical AI development means engaging robustly and transparently with these debates.
Humanities disciplines have particular expertise in navigating and guiding discussion on complex issues: this capability will be vital to the public conversation that must take place if the Plan is to succeed.
