Queensland University of Technology

Today, we find a world in which AI governance frameworks are multiplying faster than they can be reconciled. Hundreds of national strategies, sectoral regulations, and international principles now exist in parallel — overlapping in scope, competing in approach, and speaking past one another. No single framework is wrong. But without a shared architecture to orient them, the cumulative effect is fragmentation rather than governance. Smaller nations and under-resourced actors bear the greatest cost of this complexity, lacking the capacity to navigate a landscape that was never designed to be navigable. What is a solution to this regime complex and the fragmentation it is causing?

Interoperability of governance approaches

All seven thematic areas identified are important. The challenge is not choosing between them, it is understanding how they relate to one another. Safe, trustworthy AI cannot be achieved without transparency and accountability. Human rights protections depend on meaningful human oversight. Capacity-building is only effective if smaller nations can engage with governance frameworks that are designed to be accessible. Open-source models and data are governance instruments as much as technical ones. Each area is real. Each is urgent. But none of them can be fully addressed in isolation and that is precisely the problem the Dialogue must confront. The current landscape of AI governance is characterised by proliferating frameworks that address these themes in parallel, without a shared architecture to connect them. This is why interoperability of governance approaches is not simply one priority among others. It is the condition that makes progress on all the others possible at global scale.

Environmental impacts are not covered.

Organisations face a proliferating landscape of national regulations, sectoral requirements, and international principles that overlap in scope, conflict in approach, and offer no common reference point. For smaller firms, startups, and businesses in developing economies, it is a structural barrier. For larger tech conglomerates, this creates conditions for a race to the bottom that is rarely named as such. When compliance burdens are high and standards are incoherent, businesses rationally gravitate toward jurisdictions with lighter requirements, not necessarily because they oppose responsible AI, but because of the cost of this complexity. The result is regulatory arbitrage rather than genuine accountability. The opportunity cost is equally significant. Sectors such as financial services, insurance, and healthcare, where AI adoption is accelerating rapidly, could benefit enormously from clear, interoperable governance standards. Trustworthy AI is not only an ethical imperative; it is a commercial one. Businesses want to build products that can be deployed globally, that earn user trust, and that do not expose them to unpredictable regulatory liability. The current fragmentation makes all three harder.

Create a process for consolidation of existing governance efforts.

I would like to offer my current research as a direct contribution to this work. In May, I am launching a global Delphi study with experts from academia and norm-setting organisations — including the EU and UN — to design a soft governance architecture capable of consolidating existing frameworks into a coherent model. Underpinned by New Governance Theory and drawing on the precedent of the UN Guiding Principles on Business and Human Rights, the research is designed to produce a global soft governnace model based on existing norms. Its findings can be made available as they emerge.

Different stakeholders bring different and irreplaceable knowledge. Governments bring legitimacy and implementation experience. Academia brings conceptual rigour. The private sector brings the lived reality of navigating incoherent governance landscapes. Civil society represents those most exposed to AI's risks. The technical community understands what governance can and cannot realistically achieve. All are necessary. The Dialogue should resist structures that privilege any single category over others. On format, the Geneva session should prioritise depth over breadth. Rather than wide-ranging thematic panels, structured presentations by different stke holders working directly on governance would produce more actionable outcomes.

My research on developing a global soft governance model that is able to consolidate existing inititives.