Hinton warns again of major AI risks
The AI pioneer says it should be clear within ten years whether the dangers can be managed.
Listen to this article
There is no audio version yet. Request one and an AI voice will read the article aloud.
Read by an AI voice.
Geoffrey Hinton, one of the founders of modern artificial intelligence, warns that the world may not be adequately prepared for highly powerful AI systems. In an interview, he says it should become clear within ten years whether the risks can be managed.
Hinton made his remarks in an interview summarised by The Guardian. He cites possible scenarios in which AI systems are used to attack electricity grids or temporarily disrupt financial systems. These are examples of risks, not reported events that have taken place now.
According to the newspaper, the scientist stresses that despite his warnings he still hopes a catastrophe can be prevented. His central point is that AI development may move faster than the ability of governments and companies to test safety and prevent misuse.
Hinton is regarded as an influential researcher in the field of neural networks. The scenarios mentioned describe possible dangers; on their own, they do not show how likely those dangers are. When systems might reach general or superhuman capabilities, and how great the risks are, remains uncertain.
The debate touches on choices about oversight, cyber security and the speed at which new models are released. Stricter requirements can reduce risks, but may also slow research and innovation. One policy question is how much scope companies should be given to experiment and which security checks are needed before introduction.
The scenarios mentioned by Hinton also raise questions for the Netherlands about protecting energy supplies and financial systems from AI misuse. Hinton provides a prominent warning, not evidence that a disaster is imminent. The specific policy question is which risks can be made measurable and manageable before systems are deployed on a large scale.
Also show where in Europe and in the world
One story, several perspectives
What is established
- Hinton warns of possible misuse scenarios involving powerful AI systems.
- His examples are scenarios, not reported current incidents.
- There is no established date for when such systems might emerge.
Left
Arguments The risks justify robust public rules, independent safety tests and liability for companies releasing powerful systems.
Values Public safety, protection of workers and democratic control over technology.
Consequences Greater oversight could prevent serious harm, but could slow the development of useful applications.
Centre
Arguments Rules should be proportionate and risk-based: stringent requirements for critical applications, with room for research subject to clear safety standards.
Values Institutional reliability, innovation and practical implementation.
Consequences A phased approach could improve safety without bringing all AI development to a halt, but it would require continuous evaluation.
Right
Arguments Governments should focus primarily on protecting vital infrastructure and combating misuse, without broad control over technology or free development.
Values Innovation, competitiveness and limited government power.
Consequences Faster innovation could bring economic benefits, but security problems could grow if companies fail to comply with voluntary standards.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
Hinton’s statements and the examples he mentions can be found in The Guardian. The text avoids claiming that an AI catastrophe has been predicted or proven and notes the uncertainty surrounding timing and likelihood.
- confirmed Geoffrey Hinton says it should be clear within ten years whether AI risks can be safely managed. — The Guardian summarises the interview of 3 November. source
- confirmed Hinton mentions attacks on electricity grids and disruption of financial systems as possible scenarios. — The Guardian attributes these examples to Hinton. source
- uncertain There is no broad consensus on when AI will reach superhuman capabilities. — This is a broad interpretation without a single cited consensus study; the wording has therefore been kept cautious. source
Editor's note
The current statements are based on The Guardian. The text distinguishes Hinton’s scenarios and warnings from established events and scientific consensus.Sources
More on this in Dutch media
- RTL Nieuws — „kunstmatige intelligentie”
- FD — „kunstmatige intelligentie”
- Trouw — „kunstmatige intelligentie”