In July 2026, the United Nations (UN) convened the first Global Dialogue on Artificial Intelligence (AI) Governance in Geneva. It was a genuinely inclusive attempt to bring Member States, civil society, and the private sector together for the first time across four thematic clusters to discuss shared rules for AI.*
Among them, questions about how to protect and promote human rights emerged as a defining challenge, one being framed around transparency, accountability, and robust human oversight of AI. The discussion also sat squarely within global power dynamics. As Volker Türk, High Commissioner for Human Rights, said in his remarks, “AI is about power […] human rights… are a check on the exercise of power.”
Central to this discussion was the justice sector’s role as an enforcer of emerging AI regulations, a driver of accountability, a provider of remedies, and a leader in defining new jurisprudence. This recognition came from those outside of the justice sector, as well as within. Although justice was acknowledged as a key component of rights-respecting AI governance, justice actors were mostly absent from the room.
This piece examines that absence. As the global architecture of AI governance is being drawn up, with a second Global Dialogue taking place next year in New York, justice institutions cannot remain outside a conversation about a matter that will, sooner than later, fall back to them to adjudicate.
To close this gap, every AI debate underway today should include three questions: what rights should apply to AI systems, who enforces them when they are violated, and is that enforcement equally accessible to all?
The first question has occupied most Member States since the debate came to the UN during the negotiations for the Global Digital Compact in 2024. The second is where justice actors belong, and where, so far, they have not been formally invited. The third brings the conversation back to power and anchors the protection of political, economic, and social rights in effective access to justice.
