Industry Engagement and AI Aren’t Separate Conversations. Universities Should Stop Treating Them That Way.
There’s something quietly revealing about conferences where the mood is optimistic, the debates are thoughtful, and yet you leave with a sense that the most important questions still haven’t quite been answered.
That was my experience at the Quality in Postgraduate Research Conference (QPR2026). The setting was excellent, the atmosphere open and collegial, and the intent genuine: universities are trying to grapple with two pressures that now sit at the centre of postgraduate research—artificial intelligence and industry engagement. But as the conversations unfolded, I couldn’t shake the feeling that we’re still addressing these issues in parallel, rather than confronting how deeply entangled they really are.
Dr. David Beecham at the Quality in Postgraduate Research Conference
Across the conference, industry engagement, workforce mobility, and researcher skills development featured heavily. Delegates discussed a range of frameworks, including the recently released ACGR National HDR Researcher Development Framework, many drawing on the UK’s Vitae model. It was encouraging to see explicit acknowledgement of cultural knowledge, place, and neurodivergence. These are important shifts.
Yet a persistent question remained unresolved: how are core capabilities such as resilience, leadership, influence, and the management of interdisciplinary, multi-stakeholder projects across academia and industry actually taught and assessed in practice?
Having reviewed researcher training programs across the UK, Denmark, France, the US, and Canada while developing the National Industry PhD Training Program, this gap was impossible to ignore. The frameworks are improving; the pedagogy and assessment often lag behind. And judging by conversations at QPR, I wasn’t alone in that concern.
Even more striking was how little attention was paid to the professional development of current academic supervisors. Universities frequently speak about preparing the next generation of industry-engaged researchers, but far less about supporting the existing cohort of supervisors who are already being asked—sometimes pressured—to supervise industrial doctorates and engage meaningfully with industry.
The evidence here is sobering. Compagnucci and Spigarelli’s 2025 systematic review of industrial doctorates highlights supervision as a major site of complexity: candidates answer to both academic and company supervisors; responsibilities are often overlapping or unclear; and increasing workloads coupled with diminishing resources push academics toward funding acquisition at the expense of supervision itself. This is not a peripheral issue. It is a structural one.
This reality informed one of the core design decisions in the National Industry PhD Training Program: training is not just for candidates. Academic and industry supervisors are included, because effective industrial doctorates depend on capability on both sides of the partnership. Without this, ambitions for greater industry engagement rest on uneven ground.
Training provision across Australia is diverse. Some universities embed professional and research skills within the PhD; others offer them as optional. Cooperative Research Centres, Trailblazers, and ARC Training Centres play important complementary roles. Diversity here has value—industries do require different capabilities, and no single model fits all contexts.
But there are also core skills that cut across sectors: communicating across academic and industry boundaries and learning to influence rather than lead. For PhD candidates in particular, influence matters. They rarely hold formal authority, but they operate at the intersection of disciplines, organisations, and incentives.
We also need to pay closer attention to industry professionals undertaking PhDs alongside their employment. A recurring theme among Industry Researcher candidates I work with is this: they may be deeply respected experts in their industry, yet are treated as novices within academic systems. Navigating that shift requires adjustment not only from candidates, but from institutions and supervisors alike.
These questions take on added urgency in the context of national reform. The Ambitious Australia report recommends increased industry-aligned PhDs, Research Cadetships, and a coordinated national approach to research training. That coordination will matter enormously. Internationally, the direction of travel is toward consolidation: national priority-setting, shared expectations, and greater coherence in research training systems.
A coordinated Australian strategy should, at minimum, set shared expectations for core candidate and supervisor capabilities, fund supervisor development properly, enable real mobility between academia and industry, and address the significant variation in PhD assessment practices across institutions. With a more centralised R&D landscape emerging, the Australian Tertiary Education Commission is well placed to lead this work—if it chooses to engage seriously with the training challenge, not just the numbers.
Alongside industry engagement, AI unsurprisingly generated intense debate at QPR. Much of it centred on academic integrity, ethical use, and governance. These are legitimate concerns—and ones that deserve rigorous attention rather than moral panic. But overall, the conversation felt reactive rather than strategic.
There were exceptions. Sessions that asked what doctoral learning might need to become—rather than how to police new tools—offered a glimpse of a more future-oriented approach. Because the reality is simple: AI is already reshaping how knowledge is produced, and it will continue to reshape universities themselves.
In a recent fireside chat I had with economist Richard Holden, we spoke about how AI is transforming research in the Humanities and Social Sciences. Individual scholars are increasingly unable to compete with large, interdisciplinary teams using AI-enabled methods to analyse volumes of data that would once have been unthinkable. Similar shifts are already visible in institutional partnerships in the US and Australia, from Anthropic’s collaboration with Northeastern University to La Trobe’s rollout of ChatGPT Edu.
But there is another convergence we need to confront. As industry adopts AI for efficiency and R&D advantage, universities must prepare PhD candidates for research environments where project teams will include non-human agents. This is where the conversations about AI and industry engagement truly meet—and where the QPR discussions felt unfinished.
Any serious strategy for research training must grapple with how AI will transform universities as institutions and reshape how research, development, and impact occur across public, private, and not-for-profit sectors. That means equipping learners—from secondary school through to mid-career researchers—with the technical, ethical, and critical capabilities to work effectively across disciplinary and organisational boundaries in AI-augmented environments.
QPR has long played an important role in shaping doctoral education debates, and QPR2026 was no exception. The theme Transforming Graduate Research for the Future was well chosen. But the challenge now is translation.
Industry engagement and AI are not separate agendas. They are two dimensions of the same underlying question: what is postgraduate research for, who does it serve, and how do we prepare researchers for futures that will look fundamentally different from the past?
For governments, this means coordinated strategies rather than fragmented initiatives. For universities, it means investing in supervisors as seriously as candidates, and confronting the institutional shifts AI will demand. For industry, it means moving beyond transactional engagement toward genuine partnership in how research is conceived, conducted, and applied.
The conversations have started. The harder work—turning them into practice—has only just begun.
Dr. David Beecham, NIPhD Regional Director, Campus Plus