LibreOffice permits disclosed AI-assisted community code despite its no-AI message
The Document Foundation (TDF), which maintains LibreOffice, says the suite has no built-in AI features and that its core developers do not use large language models (LLMs). But several community contributions this year disclosed using AI coding assistants, prompting questions about how that fits with LibreOffice’s public “no AI” message.
Two April fixes credited Claude with helping identify problems, while a September change used GitHub Copilot to generate a class and LLM-assisted code completion for part of a screenshot feature. TDF says community contributors must disclose LLM use, submissions are reviewed by humans, and students may not use LLMs; it distinguishes development tools from AI features for users. The foundation reported 9.4 million downloads in eight weeks, though it said the reasons for the increase were unclear.
In short
- LibreOffice has no built-in AI features.
- Community contributors disclosed LLM assistance on several code changes.
- TDF requires disclosure and human review of AI-assisted submissions.
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
TDF can reasonably distinguish AI features in the product from tools contributors use while developing it: users receive no built-in AI service, while disclosed code still undergoes human review. Requiring disclosure and barring student use may help preserve accountability and learning while allowing other community contributors to benefit from assistance. A public “no AI” message can therefore describe the suite’s user-facing features and core development practices without ruling out every AI-assisted community submission.
The case against
The “no AI” message may lead users and contributors to understand that AI is not part of LibreOffice’s development, so accepting AI-assisted code can make that message seem incomplete even when the assistance is disclosed. Human review can check a contribution’s quality, but it does not by itself resolve concerns about provenance, licensing, security or whether the code reflects community standards. TDF could therefore face a case for clearer public language and consistent rules across contributors, including students.
Explained simply
LibreOffice is a free office software suite, including programs for writing documents and working with spreadsheets. It is maintained by The Document Foundation, alongside contributions from a wider community of developers.
Large language models are AI tools that can generate or suggest text and code. In software development, coding assistants can help find faults or produce parts of a program; those tools are separate from AI features built into the finished software for people to use.
The issue matters because LibreOffice’s public position says it has no built-in AI features and its core developers do not use large language models, while some community contributions have disclosed using coding assistants. The Foundation says contributors must disclose such use and that people review submitted code, so the discussion involves both its rules for contributions and how readers understand its public message.
Read the full article at the source →
Originally published by The Register as “LibreOffice’s ‘no AI’ stance meets LLM-assisted code”.