// ITS FOSS — LINUX & OPEN SOURCE
Open Source’s Anti-AI Sentiment Has Real Reasons, and So Does The Other Side
A request to ban AI-written code from ImageMagick stayed open for about eight months, until a maintainer closed and locked it this week because it had largely run its course and moved into a broader ideological debate.
Around the same time, a developer on X was telling followers to join the pushback.
Days earlier, Asahi Linux began moving its code from MIT to GPLv2 after finding that people were lifting its unreleased work and, in some extreme cases, having an LLM rewrite it to hide where it came from.
AI in open source has turned into something people pick a side on. Many have grown to dislike it, and the reasons are easy enough to see.
Others back it to varying degrees, and the more grounded among them make a case that holds up too.
With this piece, I want to go through the first group's complaints, then the louder end of the other camp, and then the careful argument for supporting AI.
The most common complaint is review burden. Rust's LLM policy announcement said a clean, tested PR used to be good evidence that its author had put in real effort and understood the change. LLMs, it says, have broken that link.
It also calls pasting reviewer comments into a chatbot and the reply back onto GitHub, a waste of time for both sides.
cURL hit the same wall, but from a security perspective. The project ended its bug bounty at the end of January after the share of reports that were real vulnerabilities dropped, and AI-generated slop became the majority.
The project's lead, Daniel Stenberg, said that the constant slop wore the team down.