AI-written contributions: help or new burden for volunteer projects?
October 5, 2026 · 12 min read
The FFmpeg maintainers tie AI security reports to maintainer burnout, Jimmy Wales considers what large language models mean for Wikipedia's future, and Guido van Rossum gives his view of Copilot. Together they show how the same tools look from the receiving end, where every contribution must be checked by a person.
The rules, numbers and steps worth keeping in front of you.
- AI output helps or burdens depending on who has to verify it.
- Generating is cheap; checking is not.
- Van Rossum: Copilot is usually slightly wrong but saves typing, and he reviews it himself.
- Van Rossum speaks as a user, not about AI submissions to Python.
- Wales: use AI to flag bias, check sources, fix grammar and summarize disputes for editors.
- Wales: a model does not know or care about the truth, so it cannot author facts.
- Wales: unchecked errors spread, as when vandalism reached a magazine.
- Kunhya: AI-generated security reports landed on volunteers with a 90-day deadline.
- Kunhya: little funding came with them, and the AI was publicized before fixes.
- Kunhya also calls Google a major open source supporter; it later sent patches.
- Kempf: maintainers feel constant reports personally, which feeds burnout.
- If you report a problem, bring the fix or the means to fix it.
- Rate severity honestly and do not publicize before a fix.
- Companies that depend on volunteer projects should fund them in proportion.
This is the short form. The full insight has every expert's reasoning, their own words and the sources.