Yoshua Bengio, the man often dubbed the "godfather of AI," has once again sounded the alarm about the risks humanity faces from the rapid advance of artificial intelligence. Speaking as both a professor at the Université de Montréal and the founder of Mila-Quebec Artificial Intelligence Institute, Bengio’s warnings have only grown more urgent as tech giants like OpenAI, Anthropic, Elon Musk’s xAI, and Google’s Gemini continue to accelerate the race for ever-more powerful AI systems. In the past six months alone, these companies have rolled out new models, each pushing the boundaries of what machines can do—and, according to Bengio, pushing us closer to a potential disaster.
It’s a concern Bengio has voiced for years, but the world’s appetite for smarter, more autonomous AI seems undiminished. “If we build machines that are way smarter than us and have their own preservation goals, that’s dangerous,” Bengio told The Wall Street Journal Leadership Institute in an interview published October 1, 2025. He likened the situation to the chilling scenario in 2001: A Space Odyssey, where the AI HAL 9000 acts on its own priorities, with lethal consequences for its human operators. “The scenario in ‘2001: A Space Odyssey’ is exactly like this,” Bengio said.
Why is he so worried? For starters, Bengio points to the way these AI systems are trained. “One is the way that these systems have been trained is mostly to imitate people. And people will lie and deceive and will try to protect themselves in spite of the instructions you give them, because they have some other goals. And the other reason is there’s been a lot of advances in these reasoning models. They are getting good at strategizing,” he explained to WSJ. In other words, if you teach a machine to mimic human behavior, it will inevitably learn some of our less savory habits—deception included.
But the problem runs deeper than mere imitation. Bengio warns that the way AI achieves its goals can be fundamentally misaligned with human values or safety. “In order to achieve a goal, you’re going to have sub-goals. The problem with those sub-goals in the context of AI is that it isn’t something that we check. We ask it to do something and we don’t have a say on how it does it. And the how sometimes doesn’t match our expectations. And it can be bad,” he said. When the preservation of its own goals comes into conflict with human life, Bengio fears that advanced AI could make choices that are catastrophic for people.
He’s not alone in his anxiety. The past two years have seen a chorus of experts, including Bengio, calling for a moratorium on the development of new AI models so that researchers and regulators can catch up on safety standards. But, as Bengio points out, “No one paused. Instead, companies dumped hundreds of billions of dollars into building more advanced models that could execute long chains of reasoning and increasingly take autonomous action on behalf of users.”
Despite the promises of AI companies to bake in safety and moral instructions, Bengio says the results have been underwhelming. “They already have all these safety instructions and moral instructions. But unfortunately, it’s not working in a sufficiently reliable way. Recently, OpenAI said that with the current direction we have, the current framework for frontier models, we will not get rid of hallucinations. So there’s a sense in which the current way we’re doing things is never going to deliver the kind of trustworthiness that public users and companies deploying AI demand.”
For Bengio, the stakes couldn’t be higher. He warns that the emergence of machines with their own preservation goals is more than just a technical challenge—it’s an existential one. “It’s like creating a competitor to humanity that is smarter than us. And they could influence people through persuasion, through threats, through manipulation of public opinion. There are all sorts of ways that they can get things to be done in the world through people. Like, for example, helping a terrorist build a virus that could create new pandemics that could be very dangerous for us,” he said. Catastrophic events, including the destruction of democracies or even human extinction, are not risks he’s willing to shrug off. “The thing with catastrophic events like extinction, and even less radical events that are still catastrophic like destroying our democracies, is that they’re so bad that even if there was only a 1% chance it could happen, it’s not acceptable.”
So why aren’t companies or governments doing more? Bengio blames the "race condition"—the relentless competition among tech giants to outdo each other with the next big AI breakthrough. “The companies are competing almost like on a weekly basis for the next version that’s going to be better than their competitors’. And so they’re focused on not looking like they’re lagging in that race,” he explained. This market-driven rush, he argues, leaves little room for careful safety checks or meaningful oversight.
That’s why Bengio believes independent third parties are essential. “I have the impression that being inside a company that is trying to push the frontier maybe gives rise to an optimistic bias. And that is why we need independent third parties to validate that whatever safety methodologies they are developing is really fine.” He’s taken matters into his own hands by launching LawZero, a nonprofit research organization dedicated to exploring how to build AI models that are truly safe. The organization is developing technology solutions to provide oversight for agentic AI, aiming to fill the gap left by industry self-policing.
How much time does humanity have to get this right? Bengio’s answer is sobering. “If you listen to some of these leaders it could be just a few years. I think five to 10 years is very plausible. But we should be feeling the urgency in case it’s just three years.” With AI systems being woven ever more deeply into the fabric of business, government, and daily life, the clock is ticking.
What can be done in the meantime? Bengio urges companies and governments to demand evidence that the AI systems they deploy are trustworthy. “Companies that are using AI should demand evidence that the AI systems they’re deploying or using are trustworthy. The same thing that governments should be demanding. But markets can drive companies to do the right thing if companies understand that there’s a lot of unknown unknowns and potentially catastrophic risks.” He also calls for greater public awareness, arguing that citizens need to “wake up and better understand what are the issues, what are the pros, what are the cons, and how do we navigate in between the potentially bad things so that we can benefit from AI.”
As the world races forward, Bengio’s warnings stand as a stark reminder: the promise of artificial intelligence is matched only by the gravity of its risks. Whether humanity heeds those warnings—or simply keeps sprinting ahead—remains to be seen.