LibreOffice's 'no AI' stance meets LLM-assisted code
Core developers avoid AI coding assistants, while community contributors face disclosure and review rules
The Document Foundation (TDF) has confirmed that LibreOffice does not employ any built-in AI features, nor do its core developers utilize large language models (LLMs). However, community contributors are subject to a different set of rules. A post on the Fediverse by user argv[-1] revealed three LibreOffice commits from this year that acknowledged LLM-assisted coding assistance.
Two commits, made in April, utilized Claude to help identify fixes for paragraph style font size and relative font sizes in Calc cell styles. In September, the Add Copy Image button to screenshot dialog credited GitHub Copilot in VS Code for generating the BitmapTransferable class, with the copyButtonHandler method written manually with LLM/code-completion assistance.
Despite the existence of LLM-generated code, TDF maintains that LibreOffice is AI-free. This stance may be in response to its former development partner Collabora, which integrated increased AI features into its web-based office suite, Collabora CODE. TDF has made the absence of built-in AI a selling point and has acknowledged a significant download increase for LibreOffice.
The foundation requires contributors to disclose LLM usage and conducts human reviews to ensure code quality. This policy seems to contrast the public no-AI message, causing frustration among some LibreOffice users.
Written by urgent.news from The Register Science's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
