As artificial intelligence rapidly reshapes classrooms around the world, much of the conversation has focused on helping students and teachers use new technologies. But according to Professor Kalervo Gulson, one of the more pressing questions is not how we use AI – it's how we govern it.
A leading expert in the governance of AI in education, Professor Gulson's research is helping governments and international organisations navigate the complex policy decisions surrounding AI adoption. His work with colleagues, nationally and internationally, has informed Australia's first public policy report on governing AI in education and is contributing to UNESCO's global work on the future of AI in education.
For Professor Gulson, the decisions being made today will shape not only the future of schools, but society itself.
While developing AI capability remains important, Professor Gulson believes governance provides the framework needed to ensure AI aligns with the fundamental values and purposes of education.
"Understanding how to use AI and building capabilities are important, but governance helps us focus on values," he says.
"AI – from today's generative AI through to emerging autonomous, multi-agent systems – has the potential to augment, and in some cases replace, humans across almost every aspect of schooling, from administration and operations to teaching, learning and student-teacher interactions."
As AI capabilities continue to accelerate, Professor Gulson argues education systems need to ask deeper questions.
"What will schooling as a human endeavour look like if AI begins undertaking many of the human-centred functions of education? Governance helps us build pathways to address these questions before the technology outpaces our ability to respond."
Understanding how to use AI and building capabilities are important, but governance helps us focus on values
Countries around the world are racing to integrate AI into education, but Professor Gulson warns that the pressure to keep pace can come at the expense of careful decision-making.
"The biggest challenge is balancing the fear of being left behind with the need to pause and think through what introducing AI across education actually means," he says.
"Governments are understandably focused on providing access to AI and building students' skills, but broader questions are often sidelined."
Those questions include how AI could change our relationship with knowledge, influence trust in schools as public institutions and contribute to environmental impacts through its rapid adoption.
While governments and international organisations have introduced a growing number of AI governance frameworks, regulations and guidelines, Professor Gulson says it remains unclear whether many of these approaches are working in practice.
"Australia can learn from international experience that focusing solely on national interests won't be enough. AI in education is a global issue that requires global cooperation."
Professor Gulson's research also examines who develops AI technologies and whose interests they ultimately serve.
Today, only a small number of global technology companies provide many of the AI systems being adopted by education providers worldwide. While collaboration between industry and education has long delivered valuable innovation, Professor Gulson says governments should not assume commercial priorities automatically align with educational ones.
"Technology companies have a history of working productively with education systems, but also of assuming that what's good for companies is good for education," he says.
"That doesn't have to be the case."
Instead, governments and institutions should evaluate AI against education's broader public purposes.
"Education supports social mobility, develops skills for the workforce and strengthens citizenship and social cohesion. We need to ask whether the AI systems we're introducing will help or hinder these goals. And if it is the latter we should be very wary of introducing AI"
Professor Gulson believes AI should be developed with education's public purpose at its core and suggests Australia should be considering what a public AI infrastructure could look like in the future.
The biggest challenge is balancing the fear of being left behind with the need to pause and think through what introducing AI across education actually means
Ensuring AI benefits everyone will require more voices in the conversation.
Professor Gulson believes decisions about AI have too often been made without those most affected by them.
"Far too often, the people most impacted by AI aren't in the room deciding whether or how it should be used."
He advocates for a collaborative "social contract" for AI in education, bringing together students, teachers, policymakers, researchers, industry and community representatives to establish shared principles for how AI should be introduced into schools.
Such an approach, he says, even if it is dealing with AI already in schools, would provide greater legitimacy and help ensure AI supports equity, transparency and the public good.
Looking ahead, Professor Gulson hopes research shifts beyond questions of access and technical capability towards understanding AI's broader social impact.
"I'd like to see us move beyond focusing only on immediate issues like skills or access to technology," he says.
"We need to understand how AI will affect individual and collective agency, and what that means for education's role in democracy."
Answering these questions will require expertise from across disciplines and sectors.
"They aren't problems that can be solved by one discipline, government or industry alone."
Through initiatives including the Centre for AI, Trust and Governance, Professor Gulson believes the University of Sydney is well positioned to bring together researchers, governments, industry and civil society to shape governance frameworks that ensure AI strengthens education while protecting its enduring public purpose.
Professor Gulson works with industry, government, education providers, technology organisations and community partners seeking to better understand and shape the role of AI in education.
Collaborative research can help develop evidence for more effective AI governance, support education systems to make informed decisions about emerging technologies, and ensure AI is aligned with the broader public purposes of education.
Organisations interested in partnering with the University of Sydney on research into AI, education, governance and technology are encouraged to get in touch to explore collaboration opportunities.
Learn more about partnership opportunities here.
