OpenAI's Patch the Planet funds open-source security fixes
OpenAI's Daybreak initiative debuts Patch the Planet, pairing AI-assisted vulnerability detection with expert review and funding for open-source maintainers.
OpenAI launched Patch the Planet this week, a Daybreak initiative that funds open-source maintainers to find and fix security vulnerabilities using AI-assisted tooling and expert validation. The program targets the chronic underfunding of open-source security work by covering both discovery and remediation costs.
Maintainers submit projects for review. OpenAI's systems scan codebases for vulnerabilities, flag candidates, and route findings to human security experts who validate exploitability and severity. Accepted patches receive funding—amounts vary by project scope and complexity—and maintainers retain full control over merge timing and disclosure.
Funding structure
Patch the Planet pays per validated fix, not per line of code. OpenAI covers the cost of the security review itself, the maintainer's time to implement the patch, and any necessary regression testing. The initiative prioritizes widely deployed libraries and frameworks where a single vulnerability can cascade across thousands of downstream projects. Funding decisions rest with Daybreak's advisory board, which includes maintainers from the Apache Software Foundation, the Python Software Foundation, and the Linux Foundation.
The program is open to any open-source project with a recognized OSI-approved license. Applications are live on the Daybreak site, and the first cohort of funded patches is expected to ship in Q3 2026.
