The Australian Academy of the Humanities has made a submission to the Joint Select Committee’s Inquiry into Artificial Intelligence, arguing that Australia’s AI future will not be decided by technical capability alone — it will be shaped by the choices we make about what AI is for, how it is designed, whose knowledge it draws on, and who benefits.
As the national body for the humanities and one of Australia’s five Learned Academies, the Academy brings expertise in the fields that explain how people, cultures, languages and institutions shape — and are shaped by — technological change. This year it established a national Humanities & AI Capability Forum to bring that expertise to bear on public policy.
Humanities led Research & Development
Our submission emphasises the unique role of Humanities-led R&D which places human, social and cultural questions at the beginning of research development, not at the end.
Examples of humanities-led R&D include the Australian Research Council Centre for Excellence for Automated Decision-Making and Society (ADM+S), the University of Sydney’s Centre for AI, Trust and Governance and the UNSW’s creation of a suite of skilled AI companions for wellbeing, developed within its Psychosocial AI program at the Big Anxiety Research Centre. All of these capability building initiatives are led by humanities researchers. These approaches genuinely bring new thinking to how Australia can navigate its AI future and determine what AI is used for, how it is designed, who benefits from it, and how widely those benefits are shared.
The submission welcomes the new Office of Artificial Intelligence within PM&C and the National AI Plan’s focus on sovereign capability and fair, inclusive benefit-sharing. But it argues these goals can’t be met without drawing systematically on humanities and arts expertise — and no policy mechanism yet exists to do that. It proposes three priorities:
- Develop a National AI Research Plan — Develop a cross-government National AI Research Plan to build and steward Australia’s sovereign capability to design, develop, adapt, evaluate and govern AI. The Plan should draw on the full spectrum of Australian research expertise, including humanities, arts and Indigenous research; support humanities-led R&D; connect research with industry and communities; and position Australia to take advantage of opportunities including the ARC’s new National Competitive Grants Program and Horizon Europe.
- Protect Australia’s knowledge and data sovereignty — Invest in the Australian knowledge, collections and data infrastructure needed to develop useful, trusted and sovereign AI. This includes digitising and curating national research and cultural collections; developing high-quality Australian cultural, linguistic and domain-specific datasets; strengthening governance and access protocols; and ensuring First Nations leadership in relation to First Nations knowledges and data.
- Build a broader national AI capability — Move beyond a narrow focus on technical AI specialists to consider the combinations of knowledge, skills and judgement Australia needs to create, use, evaluate and govern AI across the economy and society. Building on Jobs and Skills Australia’s ‘skills first’ approach, national capability planning should recognise technical, applied, strategic and governance capabilities—and the essential contribution of humanities, arts, social sciences, Indigenous knowledge and domain expertise.
Australia has a significant opportunity to shape AI in the national interest: to lift productivity, generate new industries and services, strengthen regional economies and build areas of international comparative advantage, while enriching cultural life, widening access to knowledge and services, strengthening democratic institutions and supporting human creativity and wellbeing. Realising that opportunity, and ensuring its benefits reach all Australians, will require Australia to make full use of the research, knowledge, cultural and workforce capabilities it already has. The humanities are a critical part of that national capability.
Telling the story of humanities-led R&D
The case studies below demonstrate the significant opportunities that extend from humanities-led R&D. They cross communities, political spectrums and regions, highlighting ways in which social and cultural capabilities can work closely with real people for individual, community and national benefit.
The humanities contribution is not simply to make technology ‘more ethical’. It changes what is researched, how the problem is defined, the objective/s of the research, what gets measured, and ultimately what technological systems are built and how they are deployed.
Meet Viv: humanities-led R&D extending the reach of human services through human-centred AI
About the project
Scientia Professor Jill Bennett’s FAHA work combines art, humanities, lived experience, creative practice, immersive media and AI to address a concrete social and health challenge: loneliness and anxiety among people living with cognitive impairment at home.
Viv is an AI character co-created with women living with dementia. By putting lived experience and human relationships at the centre of technological design, this kind of R&D creates possibilities for extending health, mental-health and social support beyond traditional service settings, with transformative potential for regional and remote communities where access to specialist services is constrained.
Viv is evidence of Australia’s humanities-led AI capability delivering service innovation and addressing regional inclusion.
Indigenous data governance infrastructure: governing knowledge according to cultural authority, collective benefit and community-defined priorities
About the project
Led by Laureate Professor Marcia Langton AO FASSA through the Indigenous Data Network and supported by the Australian Research Data Commons HASS and Indigenous Research Data Commons, the Improving Indigenous Research Capabilities project is developing the frameworks, skills, tools and infrastructure needed to enable Aboriginal and Torres Strait Islander communities to exercise greater authority over data concerning their lives, cultures, lands and futures.
Its 2026 Framework for the Governance of Indigenous Data was developed through three years of consultation and co-design and embeds Indigenous governance within national research infrastructure.
By bringing together Indigenous knowledge, humanities research, data science, digital infrastructure and governance, the project demonstrates that responsible AI capability requires more than technical expertise: it requires the ability to govern knowledge according to cultural authority, collective benefit and community-defined priorities.
Australian Ad Observatory: humanities-led R&D producing new infrastructure for democratic accountability in the digital age
About the project
Led by the ARC Centre of Excellence for Automated Decision Making and Society (ADM+S), the Australian Ad Observatory combines media and cultural research with computational methods, law and citizen science. Lead researchers have developed ways to make otherwise invisible platform advertising observable, including who is targeted, what messages they receive and how advertising is sequenced around users’ identities and behaviours.
Its research has exposed hidden political influence, harmful advertising and significant transparency gaps, while providing evidence for regulators, journalists and civil society. As generative AI makes political communication increasingly automated, personalised and difficult to observe, this capability is becoming essential national infrastructure for understanding and protecting democratic debate.
The Digital Inclusion Index
About the project
The Australian Digital Inclusion Index has transformed an abstract problem of ‘digital inequality’ into measurable, geographically and socially specific evidence that now guides policy, investment and service design for the non-profit sector, governments, and business. The Index tells us which Australians are best placed to take advantage of digital services, and where we need to take action to ensure no one is left behind.
The Index, together with further detailed research with First Nations communities, enables us to gauge progress towards the national Closing the Gap digital inclusion targets, which are closely related to progress in health, education, and employment.
Most recently, the Index is revealing more about the uptake of AI across Australia: how Australians are using AI, what they know about it, and what the barriers are to wider use. We aim to learn more about the critical capabilities Australians will need to maximise the benefits and mitigate the risks of AI-driven transformations.
Relational Futures: putting Indigenous sovereignty at the heart of AI
About the project
Led by Distinguished Professor Bronwyn Carlson FAHA and Dr Tamika Worrell at Macquarie University, Relational Futures examines how Aboriginal and Torres Strait Islander peoples experience and imagine AI. The Indigenous-led project identifies opportunities, including better access to services and stronger connections, alongside risks such as surveillance, bias, cultural appropriation and misuse of Indigenous data.
Relational Futures draws on the holistic and relational character of Indigenous knowledge systems to examine the full AI ecosystem, not only the technology, but its connections to Country and the environment, culture, language, sovereignty, data, community and collective wellbeing. This approach changes the questions asked of AI: not simply whether a system works, but who controls it, whose knowledge it draws upon, what relationships it creates, its impact on the environment and who benefits.
NFSA’s Bowerbird AI: domain knowledge driving technological innovation
About the project
Australia’s Galleries, Libraries, Archives and Museums (GLAM) are sites of humanities-adjacent R&D and innovation, combining technological experimentation with cultural knowledge, creativity, public service, stewardship and responsibility.
The National Film and Sound Archive (NFSA) recognised that generic speech-to-text systems were poorly suited to Australian English, accents, place names, colloquialisms and Aboriginal and Torres Strait Islander terms. Faced with hundreds of thousands of hours of audiovisual material that were difficult to search, the NFSA developed its own machine-learning transcription engine, trained to recognise Australian speech, places, people and cultural references.
Bowerbird has enabled more than 150,000 hours of material to be transcribed and linked, unlocking previously hidden stories and creating new opportunities for research, creative production, industry reuse and public engagement.
The innovation is therefore not purely technical: it combines machine learning and computational with cultural, curatorial, linguistic, archival and preservation knowledge, underpinned by a responsibility to maintain trust and create public value.
Cultivating Inclusive and Sustainable AI Policy for the Creative Sector
About the project
Australia’s role as a partner in a major Horizon-Europe funded project CULTURAI demonstrates the international R&D value of combining humanities, creative practice, AI and GLAM expertise.
Led by Associate Professor Xin Gu through Monash University’s Faculty of Arts, this €5 million Horizon Europe project brings together universities, cultural institutions, UNESCO and other international partners to investigate how AI can strengthen rather than diminish human creativity. It will generate new evidence on AI’s impact on creative labour, develop AI governance and literacy frameworks for libraries, archives and museums, and establish Living Labs where creators co-design human-centred AI tools.
The project has the potential to create new models for AI-enabled creativity, responsible GLAM innovation and cultural-sector productivity, while giving Australia a role in shaping international policy on AI, creativity and cultural diversity.
LDaCA – turning Australia's language knowledge into AI-ready national capability
About the project
The Language Data Commons of Australia (LDaCA) demonstrates how humanities-led R&D can create foundational AI infrastructure while protecting cultural knowledge and enabling new applications.
LDaCA brings together nationally significant collections of Aboriginal and Torres Strait Islander languages, Australian Englishes, migrant and Pacific languages and other language data, and provides the standards, tools and infrastructure needed to analyse them at scale. Its distinctive contribution is not simply digitisation: linguistics, humanities and Indigenous expertise are embedded in the design of the data infrastructure, access controls and governance, combining FAIR principles for research reuse with CARE principles for Indigenous data governance. This creates new national capability for language technologies and AI, while enabling research and applications in areas including translation, speech technologies, education, emergency communication and cybersecurity.
LDaCA demonstrates that humanities knowledge can be part of the foundational data and governance infrastructure on which Australia’s AI capability depends.
