The rise of AI coding tools is triggering a fundamental shift in software licensing. A recent case involving chardet demonstrates how developers are using AI to rewrite GPL-licensed libraries under more permissive terms like MIT. The maintainer reimplemented the library from scratch using only the API and test suite, achieving significant performance improvements while sidestepping the original license.

This practice has sparked heated debate within the open-source community. Original author Mark Pilgrim objects to the rewrite, viewing it as a derivative work, while the maintainer considers it a new creation. The case highlights how AI-generated code challenges traditional notions of copyright and licensing enforcement. As code generation costs plummet, developers can now recreate entire libraries from test suites alone.

The implications extend beyond individual projects. Companies like Vercel have embraced AI rewrites for some tools while objecting to others being rewritten similarly. This creates an uncertain landscape where GPL software might reemerge as MIT, proprietary code could become open source, or abandoned projects might find new life under different licenses. The fundamental question remains whether AI-generated code can even be copyrighted, potentially placing it in the public domain.