Quoting Linus Torvalds

Linus Torvalds doesn't do hedging, and he didn't start now. Responding to pushback on the Linux Media Mailing List, the Linux kernel's top-level maintainer drew a hard line: Linux is not going to be an anti-AI project, full stop. If contributors have a problem with that, he told them, they're welcome to fork the project or walk away.

What's notable here isn't just the bluntness, which is classic Torvalds, but the specific claim he's making. He's not arguing AI is perfect or that every open question about it has been settled. He explicitly carves out the economics of AI as still murky. But he's drawing a line around one particular question, whether AI tools are actually useful for real work, and declaring it closed. A year ago, he says, that might have been debatable. Today, in his view, it isn't.

Torvalds frames AI the same way he'd frame any other tool in a developer's kit, no different in kind from a compiler or a linter, just newer and more contentious. And he's got a jab ready for anyone who still doubts its usefulness: they probably haven't actually sat down and used it. That's a pointed thing to say inside a community that has plenty of engineers who have used AI tools and remain unconvinced, but it captures where Torvalds has landed personally.

The kernel has weathered plenty of contentious calls from Torvalds over three-plus decades, from C standard fights to code-of-conduct disputes. This is another one where he's using his authority as maintainer to set direction rather than build consensus. Whether the broader kernel community falls in line or keeps arguing at the margins is the part worth watching.