Generative AI is making open source development markedly faster, but it is also creating a quality and workload problem that technology leaders can no longer treat as someone else’s concern. New research from the Linux Foundation finds that code authoring time has fallen by an estimated 52% with GenAI, yet around a quarter of incoming pull requests (PRs) are judged to be low-quality “AI slop”.
The report, Navigating Open Source Software Development in the Generative AI Era: Challenges and Opportunities for Project Communities, was published by LF Research with sponsorship from Red Hat and additional support from Futurewei, OpenSSF and LF AI & Data.
Productivity gains, with a catch
For both contributors and maintainers, the research found that GenAI boosts productivity by up to 50% across all types of open source development tasks. The benefits extend beyond writing code faster. Respondents reported average time reductions across all six development activities measured, from 19% for proposal discussions to 52% for code authoring.
The gains are not uniform, however. For high-complexity tasks, 21% of respondents said they actually take more time when using GenAI, a reminder that the technology’s value depends heavily on the nature of the work.
Developers also appear to be putting the time they save to constructive use. According to the study, 66% are using productivity gains to address technical debt, 63% to develop new features and 58% for learning and upskilling.
The maintainer squeeze
The darker side of the picture is the pressure on those who review and merge code. While 59% of maintainers report growing contribution volume since the rise of GenAI, that figure climbs to 69% among large projects. At the same time, 25% of incoming PRs are estimated to be low-quality, rising to 31% among large projects.
Lower barriers to entry are widely welcomed, with 71% of maintainers agreeing that GenAI reduces them. Yet 53% disagree that GenAI has improved community trust, suggesting that easier participation is coming at a cost to confidence in what is being submitted.
For enterprises that depend on open source components, this matters. Maintainers who are overwhelmed by volume have less capacity for the careful review that underpins the security and reliability of the software supply chain.
Governance is lagging adoption
The research suggests policy has not kept pace with practice. Only 35% of projects have a formal policy permitting GenAI use, while 32% have no policy and no plans for one. Among contributors, 61% conduct detailed reviews and full functional testing of AI-generated code before submission, but disclosure of AI assistance remains inconsistent.
Human support remains important too. Some 46% of maintainers say GenAI increases the need for mentorship of new contributors, compared with 32% who say it decreases that need.
Hilary Carter, SVP of Research at the Linux Foundation, argued that projects must plan ahead rather than react. “Open source projects need to design for a growing AI-generated contributor base, not simply react to it,” said Carter. “That means setting clear expectations for GenAI, putting responsibility for testing and disclosure on contributors and giving maintainers better tools to manage incoming pull requests. If we get those fundamentals right, AI can expand participation without overburdening and increasing costs for maintainers.”
What the report recommends
To realise the most value from GenAI in open source development, the research suggests that projects should establish clear written policies for AI-assisted contributions and strengthen transparency and disclosure requirements. It also calls for contributors to test and document AI-assisted changes before submission, and for attribution practices to be standardised. More broadly, it recommends that communities treat governance, verification and trust as core infrastructure for AI-assisted development.
Industry voices
Sponsors and supporters of the research echoed the call for investment in people and process alongside tooling. “Generative AI’s impact on open source should be measured by the quality of contributions and the health of the communities behind them, alongside productivity gains,” said Mark Collier, GM of AI & Infrastructure at the Linux Foundation. “This research shows why investment in AI tools must go hand in hand with investment in maintainers, mentorship and trust. Through LF AI & Data, we’re bringing communities together to advance open source AI and help translate these gains into sustainable progress for the open source ecosystem.”
Anni Lai, Head of Open Source Ecosystem Development at Futurewei, said: “Generative AI is fundamentally changing how people participate in open source, lowering barriers to contribution while also creating new challenges for maintainers and project communities. The answer is not to slow innovation, but to build the governance, transparency and trust mechanisms that allow AI-assisted contributions to strengthen, but not overwhelm, the open source ecosystem.”
Adrianne Marcum, Chief of Staff at OpenSSF, stressed the security dimension. “AI can write code faster, but contributors still need to understand and test what they submit. This research shows why clear contribution policies and careful review matter as AI adoption grows. Keeping open source secure is a shared responsibility, and maintainers should not have to carry it alone.”
Brian Proffitt, Senior Principal Community Architect – Strategic Insights, Open Source AI Program Office at Red Hat, added: “Generative AI offers interesting potential to make software contributions far more accessible, but true productivity means little if we compromise on quality or community trust. As the volume of incoming contributions expands, project maintainers need transparent practices, dependable model lineage and consistent governance to manage this shift effectively. Ensuring a sustainable ecosystem means backing our contributors with shared standards, reliable verification and strong mentorship—guaranteeing that open source continues to thrive on human collaboration.”
For technology leaders, the findings carry a dual message. The productivity case for GenAI in software development is strong, but the benefits are being unevenly distributed, and the burden of absorbing extra volume falls on the maintainers of projects that enterprises rely on. Organisations that consume or contribute to open source would do well to set their own clear policies on AI-assisted contributions, insist on testing and disclosure from their developers, and consider supporting the maintainers whose capacity underpins the wider ecosystem.