Zig Project's Rationale for Its Strict Anti-LLM Contribution Policy

The Zig project maintains one of the strictest anti-LLM policies among major open source projects: no LLMs for issues, pull requests, or comments on the bug tracker — including translation. Users may post in their native language and rely on others' translation tools, but LLM-generated content is forbidden.

Why the Ban?

Zig Software Foundation VP of Community Loris Cro laid out the reasoning in a post titled "Contributor Poker and Zig's AI Ban." The core idea: contributors matter more than contributions.

In successful open source projects, maintainers eventually receive more PRs than they can process. Zig's approach is to accept imperfect PRs and help new contributors improve — not just to be fair, but because each contributor represents an investment. The goal of reviewing PRs is not merely to land code, but to grow new trusted contributors who become prolific over time.

LLM assistance breaks this model entirely. Even if an LLM submits a perfect PR, the time spent reviewing it does nothing to develop a new, confident, trustworthy contributor. Cro calls this "contributor poker" — quoting the card game adage "you play the person, not the cards." In contributor poker, you bet on the contributor, not on the contents of their first PR.

Context: Fork by Bun

The most prominent project written in Zig — the Bun JavaScript runtime — was acquired by Anthropic in December 2025 and makes heavy use of AI assistance. Bun operates its own fork of Zig and recently achieved a 4x performance improvement on bun compile by adding parallel semantic analysis and multiple codegen units to the LLVM backend. However, as Bun states: "We do not currently plan to upstream this, as Zig has a strict ban on LLM-authored contributions."

Cro's argument also addresses a common rebuttal: if a PR is mostly written by an LLM, why should a maintainer spend time discussing it instead of using their own LLM to solve the same problem?

Who This Is For

Open source maintainers evaluating AI contribution policies, and developers curious about the philosophical and practical arguments behind the growing number of LLM bans in open source projects.