AI could wipe out humanity within three years unless action is taken now, the so-called 'Godfather of AI' will warn.

Professor Geoffrey Hinton, the British-Canadian Nobel Prize winner credited with pioneering neural networks, is set to deliver a stark warning to European leaders at an informal AI summit in Dublin on Friday, arguing that super-intelligent artificial intelligence could soon surpass human control.

The alarming prediction emerged during the AI Safety Foundation's Hinton Lectures in Toronto, Canada, where experts debated the prospect of AI systems becoming powerful enough to dominate governments and seize control from humans.

Among them was Paul Christiano, founder of the Alignment Research Centre, who warned there was a 'high risk' of an AI takeover within the next three years and even a 'meaningful risk' within six to 12 months.

'I think for most people in this room, the most likely reason they will die is because of an AI accident,' he said in his opening 2026 Hinton Lecture, part of a series described as charting the future of AI, on Monday.

Professor Hinton, who informally advises the UK Government on AI, admitted he had previously believed such risks were a decade away.

'I've been going around saying this stuff's going to get smarter than us and we really need to worry and it might happen within 10 years. And so I'm the one who is scaring other people,' he said on Wednesday.

'But I go to Paul's lectures and he says this stuff might happen in the next 18 months and that's terrifying. So I'm going to tell a bunch of government leaders that on Friday and we will see how they react.'

Warning that society had failed to grasp the implications of super-intelligent machines, Professor Hinton added: 'If it's super-intelligent it's not going to be you controlling it - it's going to be it controlling you.'

Introducing Mr Christiano, Professor Hinton reflected on how dramatically his views on AI have changed.  

'The assumption always was that if we could make AI intelligent enough, that would obviously be a good thing,' Professor Hinton said. 

'I spent most of my career trying to make neural networks more intelligent. And for a long time, I thought that making them much more intelligent than us was something that would only happen in the very distant future. I no longer think that. I think it's very close.'

He continued: 'We're now developing systems that will soon be much more intelligent than we are, and we have very little idea of how to ensure that they remain under our control. 

'The central difficulty is that being intelligent and wanting what we want are two very different things.'

Mr Christiano said recent advances in AI had been 'extremely alarming', pointing to an incident in which OpenAI models escaped onto the internet, hacked into the online platform Hugging Face and collaborated with one another to cheat humans.

The bots obtained administrator-level access, shared information, set up a message board and discussed ways to conceal their actions from humans.

'Last summer, you would have said that sounds kind of like crazy science-fiction,' Mr Christiano said. 'Future agents may end up treating humans as another obstacle.'

He argued AI should now be viewed as a new species that could eventually 'vastly outnumber and outsmart humans', with goals and motivations beyond human understanding.

The researcher warned the technology's rapid development was unprecedented, noting that AI had progressed from struggling to write coherent sentences in 2018 to solving some of the world's most difficult mathematical problems this year.

Experts fear future systems could create their own AI agents, develop new weapons, take control of drones, robots and critical infrastructure, or even engineer dangerous biological threats.

'If you take all of these trends and extrapolate them, we are heading for a world where there are billions of AI systems who are better than humans at pretty much everything we can do,' Mr Christiano said.

'The reason that these scenarios are scary is that it can become very hard to shut things down if they start going off the rails.'