The United Nations Global Dialogue on AI Governance opened in Geneva on July 6, convening representatives from all 193 member states for two days of negotiations over how the international community should regulate artificial intelligence. The dialogue, which is distinct from the 44-member AI for Good Global Commission announced earlier this week, is a broader intergovernmental process aimed at producing a framework that could eventually form the basis of binding international standards.
According to QNA, the talks are designed to forge what organizers describe as a “fair, inclusive, and sustainable digital order”, language that reflects the tension between developed countries, which have the most advanced AI capabilities, and developing nations, which are most concerned about being excluded from AI’s economic benefits.
China arrived in Geneva with a concrete governance proposal and a clear diplomatic posture. Beijing has positioned itself as the champion of what it calls “inclusive multilateralism” in AI governance, a framework that emphasizes equal participation by all nations, technology transfer from advanced to developing countries, and each state’s right to regulate AI within its own borders without external interference.
The framing is a direct counter to the US and EU approach, which has focused on values-based governance centered on democratic principles, human rights, and restrictions on authoritarian uses of AI. China’s proposal gives it a natural constituency among the Global South, which is skeptical of governance frameworks designed primarily by Western powers.
(Related: China Joins the UN’s New AI Governance Commission as Washington Stays Out)
Washington’s Absence and Its Strategic Cost
The US government had no formal position at the Geneva dialogue. This absence is consistent with the Trump administration’s broader posture toward multilateral AI governance initiatives, which it has characterized as premature, unworkable, and potentially constraining to American technological leadership. The administration’s view, articulated in various policy documents and congressional testimony, is that AI governance should be handled through bilateral agreements and industry self-regulation rather than through UN processes that give equal voice to authoritarian governments.
The strategic cost of that posture is becoming clearer. As CSIS has noted, China’s active participation in multilateral AI governance forums gives it the opportunity to shape the norms and standards that will govern AI development globally, including in countries that are currently choosing between US and Chinese AI platforms. When a developing nation’s government is deciding which AI governance framework to adopt, the choice between a Chinese proposal that emphasizes sovereignty and inclusion and a US framework that emphasizes democratic values is not purely technical. It is a geopolitical alignment decision, and China’s presence at the table gives it influence that the US is voluntarily ceding.
The UN Scientific Panel on AI also released a preliminary report at the Geneva dialogue, providing a technical foundation for the governance discussions. The panel’s findings, which cover AI capabilities, risks, and governance gaps — are intended to inform the negotiations rather than prescribe outcomes, but they establish a shared factual baseline that makes it harder for any single country to dismiss governance concerns as speculative.
(Related: State Council Sets Agenda: AI Breakthroughs, Balanced Trade, and Green Economy)
The Governance Gap China Is Trying to Fill
China’s governance proposal in Geneva builds on a pattern evident in other international standards bodies. In telecommunications, China used its active participation in the International Telecommunication Union to shape 5G standards in ways that benefited Huawei and ZTE. In AI, the playbook is similar: participate early, propose specific technical standards, and build coalitions with developing countries that are more receptive to Chinese frameworks than to Western ones.
The specific content of China’s Geneva proposal has not been fully disclosed, but its broad outlines, as described by Chinese officials and state media ahead of the dialogue, include a call for a UN-administered AI development fund that would provide computing resources and model access to developing countries, a framework for AI sovereignty that would allow each nation to restrict or modify AI systems operating within its borders, and a set of principles for AI safety that emphasize state oversight rather than the industry self-governance model favored by the US.
Whether China’s proposal gains traction in Geneva will depend partly on how many of the 193 member states are willing to align with a framework that gives governments broad authority over AI systems. That question is not yet settled. Many developing nations share China’s concern about being excluded from AI’s benefits but are also wary of governance frameworks that could be used to justify censorship or surveillance. The Geneva dialogue is unlikely to produce binding agreements in its first session, but it will establish the negotiating positions that will shape AI governance discussions for years to come.