Chardlet dispute shows how AI will kill software licensing

"Developers who have an eye on commercial use of their open source works tend to prefer permissive licenses like MIT's because they impose fewer obligations than copyleft licenses like GPL/LGPL, which require derivative works to be distributed under the same terms."

"Licensed code, when modified, must be released under the same LGPL license. [The maintainers'] claim that it is a 'complete rewrite' is irrelevant, since they had ample exposure to the originally licensed code (i.e. this is not a 'clean room' implementation). Adding a fancy code generator into the mix does not somehow grant them any additional rights."

"Blanchard says he was in the clear to change licenses because he used AI - Anthropic's Claude is now listed as a project contributor - to make what amounts to a clean room implementation of chardet. That's essentially a rewrite done without copying the original code - though it's unclear whether Claude ingested chardet's code during training."

Dan Blanchard, maintainer of the chardet Python library, released version 7.0 under an MIT license instead of the previous LGPL. He justified this change by claiming the new version was a clean room implementation created using Anthropic's Claude AI, listing Claude as a contributor. This shift from copyleft to permissive licensing reduces obligations for commercial developers. However, Mark Pilgrim, the original creator, disputed Blanchard's right to change licenses, arguing that LGPL requires derivative works maintain the same license regardless of rewrites. The dispute centers on whether AI-generated code that may have been trained on the original codebase qualifies as a legitimate clean room implementation exempt from copyleft requirements.

Read at Theregister