Hinton Urges Brakes on Runaway AI

Geoffrey Hinton, the Nobel laureate widely described as the "godfather" of self-learning AI, urged stronger governance for artificial intelligence at the Digital World Conference (DWC), an event co-organized by the UN Research Institute for Social Development (UNRISD), reporting by UN News states. During his remarks Hinton used the metaphor, "If you ever went out with a car that had no brake, boy, you are in trouble if you go down a hill," adding, "But you're in even more trouble if there's no steering wheel," according to the UN transcript. The comments were delivered amid a concentrated week of international AI policymaking in Geneva and elsewhere, which included sessions of the Commission on Science and Technology for Development (CSTD), per UN coverage.

According to UNCTAD's *Technology and Innovation Report 2025*, cited in UN reporting, the global AI market is projected to grow from $189 billion in 2023 to $4.8 trillion by 2033. UN News also reported warnings from UNCTAD Acting Secretary-General Pedro Manuel Moreno that AI development capacity remains concentrated in a small number of economies and firms. Doreen Bogdan-Martin, Secretary-General of the International Telecommunication Union (ITU), told the same forums that generative AI adoption in the industrialized Global North is growing nearly twice as fast as in the developing Global South, a disparity she described as risking "a second great divergence," according to UN News.

Industry-pattern observations: High-level warnings from founders and senior researchers, such as Hinton, tend to increase scrutiny from policy actors and the public, especially when coupled with quantifiable market projections and UN agency commentary. Observers following the sector will note that concentration of model development, combined with rapid market expansion, typically amplifies governance friction around compute access, data sovereignty, and safety testing. Previous reporting in outlets such as Observer and Business Insider documents Hinton's consistent emphasis on existential and operational risks, including scenarios where advanced agents pursue subgoals that can lead to self-preservation behaviours; those past statements provide continuity to his current call for controls.

The event bundles three elements that accelerate policy debate: a prominent technical authority publicly warning about risk, multilateral institutions quantifying economic scale, and UN-affiliated officials highlighting equity concerns. For practitioners, this combination raises policy-related risk factors that affect model release practices, cross-border research collaboration, and procurement in regulated sectors. It also increases the likelihood that national and regional regulators will reference high-profile expert warnings when drafting safety and disclosure rules.

• •Whether national legislators or regulators cite the DWC statements or the UNCTAD projection in forthcoming bills or guidance documents.
• •Follow-up statements from major developers, and whether they disclose safety testing frameworks or coordination mechanisms in response to UN forum outcomes, as reported by trade press.
• •Multilateral initiatives or funding commitments aimed at capacity-building in the Global South, which UN officials identified as a gap during the Geneva meetings.

Editorial analysis: Practitioners should monitor regulatory texts and procurement criteria for shifts that extend beyond narrow safety tests, including requirements for transparency, third-party evaluation, and support for infrastructure and skills in lower-income countries. Such measures, when adopted, tend to impose compliance and deployment costs that affect platform choices and operations.