LF Research report: a quarter of incoming pull requests may be AI slop

LF Research, sponsored by Red Hat, surveyed 234 open source contributors and maintainers on generative AI's impact. About 25% of incoming pull requests are estimated to be low-quality AI slop, rising to 31% on projects with more than 50 contributors. 63% of large-project maintainers cite burnout as their top concern, and 64% say reviews now take longer.

• 25% of incoming PRs are AI slop, rising to 31% on large projects
• 64% of large-project maintainers say reviews now take longer
• 35% of projects have a formal AI policy; 32% have none and no plans
• 26% of projects cannot verify whether code was AI-assisted