EFF Accepts LLM Code With Human Documentation

The Electronic Frontier Foundation this week said it will accept LLM-generated code for its open-source projects — including Certbot, Privacy Badger, Boulder and Rayhunter — but requires contributors to disclose LLM use and forbids AI-generated comments or documentation, according to Alexis Hancock and Samantha Baldwin. The policy cites concerns that LLM-produced code can contain replicable bugs and impose heavy review and refactoring burdens, and asks contributors to fully understand and test submissions.