A popular Python library is now at the center of a dispute that questions how artificial intelligence tools fit into the long-established rules of open source software. The controversy emerged last week with the release of chardet 7.0, a tool for detecting text character encoding. Maintainer Dan Blanchard described the new version as a complete, ground-up rewrite built with assistance from Anthropic's Claude Code. He told The Register the project took roughly five days and resulted in a library 48 times faster than its predecessor. Crucially, Blanchard also changed the software's license from the restrictive LGPL to the more permissive MIT license, a move intended to ease the library's path into Python's standard library. The original chardet was authored by Mark Pilgrim in 2006 under the LGPL, which requires derivatives to remain open. In a GitHub comment, a user identifying as Pilgrim challenged the new release, arguing that the AI-assisted rewrite constitutes a derivative work. From this perspective, the new version would be obligated to retain the original LGPL terms, not adopt the MIT license. This situation highlights an unresolved tension. Software developers have historically used "clean room" reverse engineering to recreate functionality without copying code. Now, AI coding assistants are performing similar conceptual work at unprecedented speed, blurring the legal and ethical lines around derivative works and ownership. The chardet update is a concrete test case for how open source communities will handle this new paradigm.

Source: Ars Technica