GCC declines LLM-generated code above 15 lines, exempts test cases
The GCC steering committee adopted a policy declining any legally significant contribution that includes or derives from LLM-generated content, using the GNU Project's existing threshold of roughly 15 lines of code or text. Test cases are the one carve-out — maintainers may accept LLM-generated tests at any size — and using an LLM for research, bug hunting, or patch review stays permitted so long as none of its output lands in the patch. Contributors to one of the most widely deployed compilers now have a concrete line count separating permitted AI-assisted work from a rejected patch.
Source: lwn.net ↗
the project will decline any "legally significant contributions which include LLM-generated content or are derived from LLM-generated content"
Why this matters
- → Major open-source project sets legally defensible boundary on AI-generated code
- → Clarifies copyright risk and contributor liability for LLM-assisted patches
- → Test cases carve-out preserves AI tooling benefits while guarding core logic