Skip to content

Libreboot AI ban lasted a month before free firmware gave it up

Comic-style gloved hand soldering a laptop motherboard under a no-robots sign as glowing robot code pours in AI agents

Libreboot's AI ban lasted about a month. On November 20, the free firmware project replaced its total ban on AI-generated code with a softer "AI Slop Reduction Policy" instead. It will still discourage AI, but it no longer promises to keep all of it out.

Lead developer Leah Rowe explained the change in a post on the project's site. "I initially imposed an outright ban on AI in the Libreboot 26.11 release, but this was pure folly," she wrote. In her words, "Free software comes first, no matter who the author is, human or otherwise" in the project.

What Libreboot banned and why

Libreboot is free boot firmware, a replacement for the proprietary BIOS or UEFI on supported computers. Rowe started it in 2013. It is a distribution of coreboot with a build system that makes installation easier. It supports ThinkPads such as the X220, T440p and T480, some Dell machines and Chromebooks.

The ban first appeared in the 26.10 release on October 15. "Clankers are banned from Libreboot," the release notes said. A file in the project's code repository added that AI-generated contributions would be removed, with "no exceptions."

Rowe gave several reasons. She argued that AI deskills programmers, that the industry is a bubble, and that a few wealthy people control the technology. She also wrote that AI scraper bots had caused outages for the project, and she later shut down its public code browser.

The ban had practical effects. Libreboot removed the crabEFI component because some of its commits were AI-assisted. It also reverted a ThinkPad battery fix on the T440p that Rowe dismissed as "AI slop" in her notes. The trouble was that the code Libreboot depends on was moving the other way.

Why the AI ban could not work

Libreboot builds on coreboot, and coreboot accepts AI-assisted patches. Contributors there sign off on their code under the Developer Certificate of Origin and take responsibility for it. That means AI-assisted code was already flowing into Libreboot from upstream.

Rowe admitted this in the new policy. Coreboot, the EDK2 UEFI code and other projects Libreboot relies on all accept such patches. Forking every one of them is not realistic for a small team. In the commit that removed the old rules, she reportedly said she was never going to be able to enforce them.

She also pointed to a detection problem. She can only act when a contributor marks a patch as AI-assisted. Without that label, she wrote, she cannot prove anything. So the project replaced a promise it could not keep with a narrower one.

What the new Libreboot AI policy says

Libreboot will now "make efforts, where feasible, to reduce the influence of AI" in its code. That can mean reverting upstream patches, rewriting them or, in some cases, forking projects. The project says plainly that it will not promise to remove all AI-generated code.

AI contributions to Libreboot itself are still discouraged, and most will likely be rejected. Exceptions will be judged case by case. "I will look at the author, and judge for myself whether the work is any good," Rowe wrote. She also allows that AI can help with reverse engineering, as long as humans then rewrite the result.

The tone of the post is far from enthusiastic. Rowe still argues that AI models are built on copyrighted work, including GPL-licensed code. She closes by saying she will "attempt" to ban all AI, and says the policy is "not set in stone" for now. The dispute over how to describe the change spilled into the coverage.

How the reversal was received

Phoronix ran the news with a headline saying Libreboot "abandons" its anti-AI policy after one month. Rowe objected on Hacker News that the policy had been replaced, not abandoned, and called the article dishonest.

On Lemmy, readers were split. One praised the project for being honest about the limits. Another accused it of dropping a moral position to avoid the work of defending it. When the ban first appeared in October, Hacker News users had argued both ways. One called AI "a commercial product created from stolen labor," while others warned that strict rules drive contributors away.

Libreboot is a small project, but the same question is splitting much bigger ones.

How other open source projects handle AI code

The strict camp is shrinking but still exists. Gentoo has banned AI-generated contributions since April 2024. NetBSD treats code from language models as tainted and requires prior written approval. Asahi Linux forbids what it calls "slop generators," and Redox OS bans AI-regurgitated code. QEMU banned AI code in 2025, though its maintainers proposed easing the rule in May 2026.

Most large projects chose disclosure and responsibility instead. Fedora adopted a policy in October 2025 that makes the contributor fully accountable and recommends an Assisted-by tag. The Linux kernel added guidance for coding assistants with a similar tag, and in October 2026 its developers agreed to stop requiring it. Debian voted in August 2026 for "responsible use" of generative AI without special restrictions.

Some projects fight the side effects rather than the tools. Curl ended its bug bounty in January 2026 after a flood of fake AI-written vulnerability reports. Its lead developer, Daniel Stenberg, said the goal was to remove the incentive for "made-up lies," while still using AI tools himself. GNOME now rejects shell extensions that show clear signs of careless AI generation.

Can open source projects ban AI code

Not completely, and Libreboot shows why. A project that builds on others inherits their rules, and most big upstreams now accept AI-assisted patches. A ban also depends on contributors admitting they used AI, which no one can check reliably.

What projects can do is set a standard and enforce it at review. Libreboot's new policy is closer to that: discourage AI, reject weak patches and own the code that ships. For a firmware project that runs below the operating system, careful review matters more than where the code came from.

Sources

Portrait of Felix Laurent, author at thehype.news

Felix Laurent

News Writer at thehype. Felix is a journalist with experience in marketing. He covers artificial intelligence and emerging technologies, focusing on the ideas, innovations, and trends shaping the future of tech.

More articles by Felix Laurent →
On air · radio.thehype.news ↗ AI news radio — 24/7