Linux Foundation survey finds faster AI coding alongside low-quality contributions

Writing open-source code with AI may take less time, but accepting the resulting contributions creates a different challenge. A Linux Foundation Research survey published October 20, 2026, puts respondents’ estimated code-authoring time savings at 52%. Maintainers, meanwhile, estimated that 25% of incoming pull requests—proposed code changes—were low-quality “AI slop.”

The report, Navigating Open Source Software Development in the Generative AI Era: Challenges and Opportunities for Project Communities, examines both contributors and maintainers. Its findings describe their estimates and experiences, rather than an independently timed test of AI tools. The foundation’s announcement does not give the survey’s sample size or field dates, limiting how precisely readers can assess its reach.

Time savings depend on the task

AI assistance reaches beyond drafting code. Respondents estimated average time reductions across all six development activities the survey measured. Proposal discussions, for example, showed a smaller estimated reduction of 19%. That spread is important: the largest coding figure should not be read as an equal saving across every part of maintaining an open-source project.

Complexity also changes the picture. For high-complexity tasks, 21% of respondents said using generative AI took more time. The survey therefore captures both substantial reported savings and cases where assistance added work, rather than a uniform improvement for every developer or assignment.

Respondents also described where saved time goes. Sixty-six percent reported using productivity gains to address technical debt, or accumulated software cleanup work. Developing new features followed at 63%, while 58% reported learning and building skills. Those answers describe several uses for the gains, not simply a faster route to submitting more code.

69%Large-project maintainers reporting contribution growth

Among large projects, 69% of maintainers reported growing contribution volume since the rise of generative AI.

31%Estimated low-quality AI pull requests at large projects

Maintainers’ estimate of low-quality AI submissions rose to 31% for large projects.

Easier entry does not mean greater trust

Across projects, 59% of maintainers reported contribution growth since the rise of generative AI. The survey pairs that increase with a participation benefit: 71% agreed that AI lowers barriers to entering open source. But 53% disagreed that it had improved community trust. Easier participation and stronger relationships are distinct outcomes in these responses.

Nor did maintainers generally describe AI as a replacement for helping newcomers. Forty-six percent said it increases the need for mentorship, compared with 32% who said it reduces that need. That finding sits alongside the report’s call for continued human support as AI makes contributing more accessible.

Who checks the code before submission?

Formal rules are far from universal. Only 35% of projects had a formal policy permitting generative AI use; 32% had no policy and no plans to create one. Among contributors, 61% reported detailed reviews and full functional testing of AI-generated code before submission. Disclosure of AI assistance remained inconsistent, according to the foundation.

The report’s recommendations put responsibility on contributors as well as project leaders. They call for checks before a proposed change reaches maintainers, alongside clearer shared expectations:

• Write clear policies for AI-assisted contributions and strengthen disclosure requirements.
• Require contributors to test and document AI-assisted changes before submitting them.
• Standardize attribution practices and continue investing in human mentorship.

Linux Foundation Research published the study with sponsorship from Red Hat and additional support from Futurewei, OpenSSF and LF AI & Data.