Linux creator Linus Torvalds tells AI haters to walk away from Linux, or go fork it

Linus Torvalds Refreshes Stance on AI in Linux Development

Creator of the Linux kernel, Linus Torvalds, has refreshed his stance on the use of AI in software development with a strongly worded statement posted on the lore.kernel.org mailing list archive service. Traditionally a staunch guardian of human-centric code quality and previously somewhat dismissive of AI, Torvalds has now pivoted to come out flags-waving, ready to support its use across the Linux codebase.

Linux is not one of those anti-AI projects

"I realize that some people really dislike AI, but this is an area where I'm willing to absolutely put my foot down as the top-level maintainer," writes Torvalds. "Linux is not one of those anti-AI projects, and if somebody has issues with that, they can do the open-source thing and fork it. Or just walk away."

The famously opinionated Finnish-American confirmed that he regards AI as a tool, "just like other tools we [software engineers] use," and further agreed that "it's clearly a useful one" today. With some balance, Torvalds also conceded that, as of now, it's "tough to know what the AI economy will look like" in the end.

More on Linus Torvalds

• Why Linux creator Linus Torvalds gets angry hearing "99% of code is AI" (2026)
• Linus Torvalds gets candid about Windows, workflows, and AI (2026)
• Linus Torvalds reflects on 20 years of Git (2025)
• Linus Torvalds on security, AI, open source, and trust (2024)
• Linus Torvalds on community, Rust, and Linux's longevity (2021)

Pivoting again, Torvalds writes that he feels AI can "also be a somewhat painful tool," both from the impact it has on maintainer workloads and from the fact that "it keeps finding embarrassing bugs" standpoint.

"But the solution is not to put your head in the sand and sing 'La La La, I can't hear you' at the top of your voice, like some people seem to do," wrote Torvalds. "The solution is to make sure those LLM tools help maintainers instead of just causing them pain. There's no question on that side."

"I realize that some people really dislike AI, but this is an area where I'm willing to absolutely put my foot down as the top-level maintainer."

"We're not forcing anybody to use it, but I will very loudly ignore people who try to argue against other people from using it," he added.

The twist is that, as Torvalds himself conceded, his opinion on the use of AI may not have been a clear yes even just a year ago, even if it is no longer in question for him today. Speaking on the video storytelling website TFIR in October 2024, Torvalds was considerably less upbeat about AI.

Torvalds in 2024: AI will be useful in five years from now

"I think AI is really interesting and I think it is going to change the world, and at the same time I hate the hypecycle so much that I really don't to go there, so my approach to AI right now is I will basically ignore it because I think the whole tech industry around AI is in a very bad position and it's 90% marketing and 10% reality," said Torvalds, at timemarker 37:57 mins.

"In five years, things will change, and at that point we'll see what (of the) AI is getting used every day for real workloads instead of just ChatGPT [being used to] make great, like, uh, demonstrations," he added.

It's actually 21 months or 637 days and counting since Torvalds voiced his anti-AI hypecycle diatribe, so that's less than five years.

"I don't really care whether AI helped write the code. I care that the person submitting it can explain their intent, why it's needed and why it fits the project."

In fairness to the chief Penguin himself, enterprise software industry headlines in mid to late 2024 were full of Kubernetes repatriation analysis, SAP's 25% surge in Q3 cloud revenues, and stories around the use of Google's preview of Memorystore for Valkey. Anthropic was at Claude 3.5 Sonnet, and Microsoft launched autonomous AI agents within Copilot Studio, fending off the challenge from Salesforce. Perhaps those really were different times.

The AI-code-quality debate is over

Luke Hinds, co-founder and CEO of open source developer security company nolabs, tells The New Stack that "Torvalds is bang on the money" and that we shouldn't be banning AI tools or putting our heads in the sand - but we need to insist humans stay responsible for thinking.

"The AI-code-quality debate is over, and honestly it was always a distraction," Hinds says. "The question that matters in open source hasn't changed: does a human stand behind this contribution and do they understand it well enough to explain it? I don't really care whether AI helped write the code. I care that the person submitting it can explain their intent, why it's needed, and why it fits the project. I want evidence of judgment, not just output."

"Torvalds is bang on the money."

The failure mode Hinds rejects is the one he explains as the one that maintainers are already seeing i.e. a pull request where an AI code agent has generated a 2,000-line comment of impressive-sounding jargon to justify a change nobody clearly reasoned through.

Mike McGrath, vice president, core platforms at Red Hat, tells The New Stack that his organization has been "very public about our view of AI-assisted development" for working software engineering teams.

"AI-assisted development is an incredibly powerful component, if not a collaborator, for solving complex engineering challenges and driving open innovation forward," McGrath says. "Human guidance and guardrails remain crucial to the sustainable use of AI in open source, as we can never lose sight of the fact that AI is a tool to reach an outcome, not the outcome itself. We want AI to be useful for everyone, and look forward to working with kernel.org maintainers in shaping future harnesses and LLMs to be open and effective for all."

Can we second-guess Torvald's next opinion?

While it might appear foolish to attempt to predict Torvald's next opinion swing, there's actually meritocracy-based methodology to his madness. He rarely stays completely anchored to one stance on any given technology, instead favoring engineering pragmatism based on what works, what doesn't, and what might work in the future. His argument might be: ideological consistency is nice, but ability-based progression and advancement based on objective open community evaluation is better.