March 4, 2026
License wars go brrr
Relicensing with AI-Assisted Rewrite
AI “rewrite” flips a popular Python tool’s license — commenters cry foul, cheer, and meme
TLDR: A major Python library was AI-rewritten and switched from share-your-changes (LGPL) to do-what-you-want (MIT), igniting legal and ethical chaos. Commenters clash over whether the code is still LGPL, whether AI output is public domain, and whether this nukes copyleft entirely.
Open-source folks woke up to drama: the maintainers of the widely-used Python library chardet say they used Claude Code (an AI assistant) to “rewrite” the whole thing and switch the license from LGPL (a share-your-changes license) to MIT (a do-what-you-want license). The original author called it a potential violation, arguing a real “clean room” rebuild requires two separate teams and zero exposure to the old code. The community promptly split into camps, and the takes are spicy.
One camp is screaming, “It’s still LGPL, you can’t just slap MIT on it!” — hardliners insist relicensing needs consent from all contributors. Another camp gleefully declares “AI outputs are public domain!”, waving at a recent Supreme Court move that reinforced a “human authorship” requirement and tossing fuel on the chaos. Then the nihilists roll in with: if AI can’t own copyright, is the new code license even meaningful?
Cue the memes: users joked we’ll see “Linux, now MIT” in a year, and asked if the sacred “clean room” idea even has real legal teeth. A few dream of an AI-shaped end-run around copyleft, while others warn it would nuke open-source norms. It’s part courtroom thriller, part Reddit roast, with the future of licenses on the line.
Key Points
- •chardet maintainers used Claude Code to rewrite the library and released v7.0.0 under the MIT License.
- •The original author, a2mark, alleges the rewrite may violate GPL/LGPL due to maintainers’ exposure to the original code.
- •The article explains that a traditional clean-room rewrite requires separation of specification and implementation teams.
- •It cites a U.S. Supreme Court move on AI authorship as creating a paradox: ownership, derivative status, and possible public domain outcomes for AI-generated code.
- •The article warns that accepting AI-based rewrites for relicensing could undermine Copyleft, enabling permissive relicensing of GPL projects.