Regulatory Interoperability for AI: where is Alcibiades’ dog tail?

When Alcibiades cut off his dog’s tail, he knew what he was doing. Athens would gossip about the cruelty and ignore the war. Politics, he understood, is often about distraction. Artificial intelligence regulation may have acquired its own dog tail.

The AI-regulation debate is too often consumed by grand gestures: calls for global AI treaties and declarations; sweeping principles and landmark conferences; and ever longer lists of values, codes of practice and “global firsts”. Meanwhile, AI systems move across borders with ease, while regulation (and risk-based analyses) do not. Models are trained in one jurisdiction, deployed in another, and governed—fragmentedly—by many (or none). Specific-purpose systems sometimes face jurisdictional barriers and some others do not, without any kind of structural logic. Courts, regulators, standards bodies and procurement rules all seem to tug at some uses. No one is in charge. Everyone claims concern.

Into this disorder steps “regulatory interoperability”. The words reassure. The concept suggests harmony without sacrifice. But too often it functions like Alcibiades’ dog: something eye‑catching that diverts attention from harder questions. What, precisely, must interoperate? Laws? Institutions? Evidence? Decisions? Or merely aspirations?

Technical interoperability we are familiar with. Regulatory interoperability we are not. It is not harmonisation in disguise, nor deregulation by another name. It may be a strategy for a world that will not agree on first bests, but we are not sure how that looks. It requires design, not slogans. Translation, not consensus. Without that work, policymakers will keep debating the tail—while the dog runs without direction.

In this seminar José-Miguel Bello y Villarino will share his observations about this process and where it seems to be heading. He will present how he sees regulatory interoperability to be distinct or similar to other related concepts and where it is another byword for misdirection or, even worse, hiding attemps to reinvent the wheel. He will engage with (and criticise) previous work, to dispiritedly offer limited (if any) solutions. He will discuss how the term has been used in policy documents, just to illustrate its operational limitations. And yet, he hopes something insightful would come out of this research agenda and seminar.

Bio: José‑Miguel Bello y Villarino is a Senior Research Fellow at the University of Sydney Law School and an Australian Research Council Early Career Industry Fellow (2025–27), affiliated with the ARC Automated Decision-Making and Society Centre of Excellence and the University of Sydney Centre for AI, Trust and Governance. He is Director of Research Engagement at the Sydney Law School, where he co‑teaches advanced Law and Technology, together with an interdisciplinary course at the Faculty of Engineering. He is a member of Spain’s Diplomatic Corps (on leave), with prior experience at the European Union (Commission and Council). His ARC fellowship, in partnership with the NSW Independent Commission Against Corruption, develops governance frameworks for using AI and automation to combat systemic corruption in the public sector. He has held visiting appointments at the EUI School of Transnational Governance and Harvard Law School and is a graduate from Santiago de Compostela (Political Sciences and Law), College of Europe (Law), NYU School of Law, the Spanish Diplomatic School and the University of Sydney.

This seminar will operate on a first-come, first-served basis.