# GitHub gives maintainers new controls as AI-generated pull requests overwhelm open-source projects > Source: > Published: 2026-09-01 16:31:08+00:00 # GitHub gives maintainers new controls as AI-generated pull requests overwhelm open-source projects The platform's new repository controls let maintainers disable or limit external pull requests after a surge in low-quality AI submissions pushed monthly PR volume to 90 million. Open-source software development has a spam problem, and it’s wearing a trench coat made of AI-generated code. GitHub has rolled out new repository-level controls that let project maintainers disable pull requests entirely, restrict submissions to existing collaborators, or cap the number of concurrent open PRs from outside contributors. The changes, which went live on August 27, 2026, are GitHub’s most aggressive response yet to a flood of low-quality, often AI-generated contributions that drain maintainer time and pollute project histories. ## The anatomy of AI slop GitHub’s own data tells the story in stark terms. Between early 2023 and March 2026, the platform’s monthly merged PR volume surged from roughly 25 million to 90 million. That’s a 260% increase in three years, and a meaningful chunk of it appears to be machine-generated noise rather than genuine contribution. The submissions in question tend to share a few telltale traits. They ignore project contribution guidelines. They introduce changes that are cosmetically different but functionally identical to existing code. They fix “problems” that don’t exist. And when maintainers leave review comments asking for revisions, the PRs are often abandoned entirely. One particularly illustrative case surfaced in April 2026, when an engineer using an autonomous coding agent submitted more than 130 pull requests across multiple repositories in just 72 hours. GitHub suspended the account. ## How the new controls work The new features give repository owners three distinct levers to pull. First, they can disable pull requests on a repository altogether. Second, they can limit PR submissions to current project collaborators. Third, they can set a cap on how many concurrent open PRs any single external contributor can have at once. GitHub product manager Camilla Moraes first initiated internal discussions about these tools back in February 2026, meaning the platform spent roughly six months developing and testing the features before public release. The controls are opt-in. Individual maintainers get to decide what level of openness works for their projects. ## The policy landscape taking shape A study examining 1,000 GitHub repositories found that 118 had already established formal AI contribution policies. Of those, 78% permitted AI-assisted work under certain conditions, things like requiring disclosure of AI tool usage, mandating human review of AI-generated code, or insisting that contributors demonstrate understanding of the changes they’re submitting. ## What this means for open-source development For GitHub as a business, the situation creates an awkward dynamic. The company’s Copilot product is one of the most commercially successful AI coding assistants on the market, and the broader ecosystem of AI agents that Copilot helped popularize is directly responsible for the PR flood that GitHub now has to manage. The platform is simultaneously selling the firehose and building the flood wall. **Disclosure:** This article was edited by Editorial Team. For more information on how we create and review content, see our [Editorial Policy](https://cryptobriefing.com/editorial-policy/).