Why Zig bans LLM contributions: contributors matter more than code

The Zig project enforces one of the strictest anti-LLM policies in major open source: no AI-generated issues, pull requests, or bug tracker comments, including translation. The policy is notable enough that Bun, the Zig-based JavaScript runtime acquired by Anthropic in December 2025, refuses to upstream a 4x compile-time performance improvement to Zig because the work was AI-assisted.

Zig Software Foundation VP of Community Loris Cro frames the rationale as “contributor poker” — you bet on the person, not the cards. Successful open source projects eventually receive more PRs than maintainers can process. Rather than gate-keeping for ROI, Zig invests review time in growing trusted long-term contributors. LLM-authored PRs break that economic model: even a technically perfect AI-generated submission yields no human contributor for the project to develop.

The policy implicitly answers a question increasingly raised across open source: if a maintainer is reviewing a PR an LLM wrote, why not skip the review loop and have their own LLM produce the patch directly? Zig’s stance reframes contribution review as relationship-building rather than code acquisition, making the ban an organizational design choice rather than a technical or ideological one.