The United Nations has consolidated a number of its concerns about artificial intelligence into a single document, and the headline finding is stark: AI capabilities are accelerating faster than any government’s ability to understand, test, or regulate them. The warning arrives as delegates gather in Geneva for the opening of the UN Global Dialogue on AI Governance, and it lands into a policy landscape where the EU’s AI Act remains one of the few binding frameworks anywhere in force.
The document behind the warning is a preliminary report from the UN’s Independent International Scientific Panel on AI, released on 1 July and billed as the first comprehensive global assessment of the technology. Its central claim is a gap: between what AI systems can now do and the scientific understanding needed to govern them. Regulation is lagging, the panel argues, but so is the foundational research that policymakers would need to write good rules in the first place.
The Core Warning: A Structural Mismatch
Secretary-General António Guterres delivered the message in plainer terms. “The more AI advances without shared rules, the less say governments and people will have in the outcome,” he told reporters, before reducing his advice to governments to two words: “Do not wait.” He returned to the theme of comprehension more than once. “The world cannot govern what it cannot understand,” he said, adding that “the potential is great, but the risks are real, and the cost of waiting is rising.”
That framing—governance chasing an object it cannot yet measure—is what gives the report its force. The panel is not primarily warning about any single catastrophic scenario. It is warning about a structural mismatch, in which the pace of capability gains outruns the slower work of evaluation, standard-setting, and law. This is a familiar complaint among researchers who study AI governance, but it carries new weight given the authority of the UN behind it.
The report draws on contributions from hundreds of experts across disciplines, including computer science, ethics, economics, and international law. It attempts to map the current state of AI capabilities—from large language models to autonomous systems—against the existing governance infrastructure. The conclusion is that the gap is widening. While AI systems are being deployed at scale in areas like healthcare, finance, and criminal justice, the tools to assess their safety, fairness, and reliability remain underdeveloped. The panel calls for a significant increase in investment in AI safety research and for the creation of global standards for transparency and accountability.
Fragmented Regulatory Efforts Worldwide
The obvious objection to the panel’s warning is that governments are not doing nothing. The European Union has a risk-based rulebook in force, the AI Act, which categorises applications by their potential harm and imposes corresponding obligations. However, implementation across member states remains uneven, and the Act’s full effects will not be felt for years. China has moved to restrict humanlike AI agents, forcing changes to consumer products already on the market. The United States, by contrast, has struggled to produce durable federal rules at all, a vacuum that critics say leaves the country poorly placed to regulate the industry it largely hosts.
The panel’s point is that these efforts are fragmented, and that fragmentation is itself a risk. Without a coordinated international framework, companies can shop for jurisdictions with the weakest rules, and governments lack a common baseline for evaluating risks. The report notes that even where domestic regulations exist, they often address narrow use cases and fail to keep pace with the rapid evolution of the technology. For example, rules designed for earlier generations of AI may not apply to generative models capable of producing convincing text, images, and video.
Furthermore, the report highlights the absence of binding international agreements on AI governance. While organisations like the OECD and UNESCO have issued principles and recommendations, these are voluntary and lack enforcement mechanisms. The UN’s own efforts, including the Global Dialogue on AI Governance, are intended to build momentum toward a more formal treaty, but progress has been slow. The panel argues that the window of opportunity to shape AI’s trajectory is closing, and that delay could lock in a system that benefits the few at the expense of the many.
Equity and the Concentration of Power
There is also an equity argument threaded through the assessment. The experts caution that if the technology remains concentrated in a handful of firms and countries, the result could widen global inequality rather than narrow it. Access to compute, data, and talent is not evenly distributed, and neither is the capacity to govern. The report points out that low- and middle-income countries are often excluded from AI development and governance discussions, even as they are disproportionately affected by its consequences—for example, through labour displacement or biased systems deployed by external actors.
The panel calls for mechanisms to ensure equitable participation in AI governance, including capacity-building programs and technology transfer. It also urges governments to invest in public AI research and open-source initiatives to reduce dependence on a few powerful corporations. Without such measures, the report warns, AI could exacerbate existing inequalities and create new forms of digital colonialism.
The Advisory Model: Lessons from Climate Science
What the report does not do is prescribe a specific institution or treaty. It feeds instead into the Geneva dialogue, which is meant to be the beginning of a process rather than a decision point. The UN has been careful to frame the panel as advisory, a scientific body modelled loosely on the climate assessments that inform intergovernmental negotiations without dictating them. Whether that model can move at the speed the report itself describes is the open question.
Intergovernmental processes are deliberate by design, and the panel’s core finding is that AI is not. The climate parallel is instructive in both directions: the assessments have produced a shared body of evidence, but decades of them have not guaranteed decisive action. The panel is betting that a common scientific baseline is still worth having, even when the politics lag behind it. In the case of AI, the stakes are arguably higher because the technology evolves on a timescale of months, not decades.
The report recommends the establishment of a permanent scientific advisory mechanism to provide ongoing assessments and early warnings. It also suggests creating a global observatory to track AI developments and their societal impacts. These recommendations echo proposals from civil society groups, but they face significant political and financial hurdles. The UN itself must rely on voluntary contributions from member states to fund the panel’s work, which raises questions about sustainability and independence.
The Geneva Dialogue and Next Steps
The Global Dialogue on AI Governance, which opened in Geneva on the same day the report was released, is intended to build on the findings and chart a path forward. The event brings together government officials, industry leaders, researchers, and civil society representatives for a series of panel discussions and workshops. The agenda includes topics such as risk assessment, international coordination, and the protection of human rights in the age of AI.
While the dialogue is not expected to produce binding commitments, the UN hopes it will generate political will and concrete proposals for the next phase of work. One possible outcome is a resolution at the UN General Assembly calling for a global framework on AI, which could pave the way for negotiations on a treaty. However, such a process would take years, and the panel’s report makes clear that the technology will not wait.
In the meantime, the report urges governments to take immediate steps, including establishing domestic oversight bodies, requiring mandatory reporting from AI developers, and investing in education and public awareness. It also calls on tech companies to adopt responsible practices voluntarily, such as conducting pre-deployment testing and sharing safety data with independent researchers. The message is that even without a global agreement, there is much that can be done now.
The UN’s intervention comes at a critical juncture. The capabilities of AI systems continue to improve at a rapid pace, driven by advances in hardware, algorithms, and data availability. The potential benefits are enormous, from medical breakthroughs to climate modelling, but so are the risks. The report warns that without better governance, AI could be used to amplify disinformation, automate discrimination, or create new forms of surveillance and control. It also notes the dual-use nature of many AI technologies, which can serve both beneficial and harmful purposes depending on their application.
Ultimately, the report is a call to action that echoes earlier warnings from other international bodies and expert groups. What distinguishes it is the breadth of its scope and the authority of its sponsor. The UN hopes that by assembling a global scientific consensus, it can create the political pressure needed to overcome inertia and resistance. Whether that strategy succeeds remains to be seen, but the report leaves no doubt about the urgency of the challenge.