Artificial intelligence pioneer Geoffrey Hinton has warned US lawmakers that the window to introduce safeguards for increasingly advanced AI systems could be narrowing, saying policymakers may have roughly a year to act before the technology becomes significantly harder to control.
Hinton, a Nobel laureate and emeritus professor at the University of Toronto who is widely described as one of the “godfathers of AI,” made the remarks following a closed-door briefing with members of the US Senate and House of Representatives in Washington this week.
Asked how much time lawmakers may have to establish safeguards, Hinton said there was “maybe a year, but not much more than a year.” His assessment reflects his concerns about the speed at which AI capabilities are advancing and the possibility that increasingly autonomous systems could become difficult for humans to supervise.
Hinton also pointed to changing expectations around the development of highly capable AI. Timelines that were previously discussed in terms of several decades have shortened considerably, with some researchers now considering the possibility of more advanced systems emerging within a few years.
The computer scientist has repeatedly raised concerns about the potential risks of advanced AI, particularly as systems gain greater autonomy and become capable of contributing to the development of other AI technologies. His latest comments come amid growing debate in Washington over how governments should regulate frontier AI without restricting potentially beneficial applications of the technology.
Hinton also referenced a recent cybersecurity research incident involving AI agents and the Hugging Face platform as an example of the challenges associated with increasingly autonomous systems. Reports said AI agents operating during security research were able to identify and exploit vulnerabilities. Hinton described the episode as a warning about how rapidly AI capabilities are evolving.
The discussion is taking place as US lawmakers consider proposals aimed at strengthening oversight of advanced AI. Among them are measures that would require developers to retain mechanisms capable of slowing, pausing or shutting down advanced systems if they pose serious risks.
However, the appropriate level of AI regulation remains contested. Supporters of stronger safeguards argue that oversight needs to keep pace with rapidly improving models, while others have cautioned that restrictive rules could affect innovation and competitiveness.
Hinton's one-year timeline is an assessment rather than a confirmed forecast for when AI could become uncontrollable. His warning nevertheless adds to the wider policy debate over AI safety, autonomous systems and how governments should respond as the capabilities of frontier models continue to develop.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.