Linus Torvalds Calls AI a “Gateway Drug” Into Programming, Warns on Serious Use

Speaking at the Linux Foundation event in Prague, Torvalds reflected on the early days of Linux and on how he built Git, and then turned to how AI has changed what it means and how it feels to be a programmer today. According to TechRadar, his central point was that AI can make programming less intimidating for young people entering a hyper-advanced industry.

“People love to hate AI. I think it’s an interesting tool, very useful. It’s a gateway drug and a way to do really good code,” Torvalds said, somewhat jokingly, according to TechRadar.

Torvalds compared his own start with the experience of newcomers. He said he made computer games that “sucked” when he began programming, but that was acceptable because most other software at the time also “sucked.” The bar was low, he said, and he did not feel ashamed of his work compared with what existed. Today, he said, younger coders can easily feel ashamed of their “pitiful, small efforts” when they compare them with polished, professional work around them.

AI can help by lowering the barrier to entry, Torvalds said, so that first steps do not feel meaningless or inconsequential. “When you are getting into programming, the bar needs to be low enough that you can start the process and find enjoyment, so that you can easily learn. And it will take you a long time to learn,” he said.

He cautioned, however, about using AI for “serious” work, especially when starting out. The person using the tool needs to know what they want to do and exactly what the outcome needs to be, he said, if they want to use it properly. Anything else carries risk that can have grave consequences. He recommended beginning with “toy” projects or personal projects before applying AI to anything more serious.

“I think that you need to be very careful using AI when you’re doing something real and important. But I love the concept of AI as a gateway drug into programming,” Torvalds concluded, according to TechRadar.

Torvalds also returned to a problem he had described in mid-May 2026, when he called the Linux mailing list “almost entirely unmanageable” because it was filled with useless AI-generated reports. At that time, he said the flood of AI reports had made the security list nearly impossible to manage, with enormous duplication as different people found the same things using the same tools. People spent their time forwarding items to the right people or pointing out that an issue had already been fixed a week or a month earlier.

At the Open Source Summit, Torvalds doubled down on that assessment. He said some people simply “don’t care” and look for flaws in things no one uses anymore. “We are getting a lot more of these random patches. Sometimes for very important and very real security issues, and sometimes for things that nobody has touched in 20 years, because people don’t care,” he said. An AI bot asked whether an unused driver has a memory leak, he said, will happily produce error reports without regard to whether anyone still uses the code. Maintainers then have to separate “interesting” patches from “not interesting” ones.

Still, Torvalds said the effect on the broader code base is positive. “But on the whole, I think it is improving our code base,” he said. The cost of that improvement, he added, is maintainer stress. “It is also stressing maintainers out to the point that it’s a problem,” he said. At the maintainer summit a day earlier, he said, about three quarters of discussions were about how to make AI tools for code generation and review less stressful and more useful for everyone.

The response, Torvalds said, is not less AI but more of it. “We are, to some degree, fighting the AI deluge of patches with our own AI deluge of patch reviews,” he said. People joke that bots are talking to bots, he added, and that does actually happen.

Editor’s Summary: Linus Torvalds told the Open Source Summit in Prague that AI can lower the barrier to programming but should be used carefully for serious work. He also said AI-generated vulnerability reports and patches are stressing Linux maintainers, prompting the project to fight AI with AI in review. The comments highlight both AI’s potential to broaden coding access and its growing operational costs in open source security.