AI cuts coding time, but it's also filling open source's inbox with crap
theregister ● Covered by 2 sources
AI is speeding up coding for open source devs, but it's also flooding maintainers with junk. A Linux Foundation survey says a quarter of incoming pull requests may be AI slop.
Based on reporting by theregister — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Generative AI is shaving time off open source coding, but it is also handing maintainers a bigger pile of low-quality work. That’s the sharpest reading of a Linux Foundation survey, which found respondents estimating that 25 percent of incoming pull requests are “AI slop.”
The upside is real. Almost three-quarters of maintainers, 71 percent, said GenAI has made it easier to contribute. Respondents estimated that code-authoring time is down by 52 percent on average. But the tools are not a universal win: for complex tasks, 21 percent said the work actually takes longer.
The trust picture is messier. Just over half of respondents, 53 percent, disagreed that GenAI has improved community trust. And the governance around all this is still shaky. Only 35 percent of projects have a formal policy allowing GenAI use, while 32 percent have no policy and no plans for one.
That leaves projects trying to cope as the volume rises. Among large projects, the estimated share of incoming pull requests classified as AI slop jumps to 31 percent. The problem isn’t limited to code changes either. The curl project said this week it will move low and medium severity flaws into public view, because its security team is nearly overwhelmed by AI-assisted reports.
Linux Foundation SVP of Research Hilary Carter said projects need to set clear expectations, put testing and disclosure duties on contributors, and give maintainers better tools. That lines up with Linus Torvalds’ stance: more open to the tools, still annoyed by the junk they can generate. Open source wanted more contributors. It got them. It also got the inbox work that comes with them.
My take — AI-written commentary, not fact-checked reporting
Open source doesn’t need more AI cheerleading; it needs spam filters with better manners. The industry keeps acting surprised that making contribution cheaper also makes low-effort noise cheaper. Clear rules and better triage tools are the boring answer, which is usually the right one.
Read more about this at: theregister
