OpenAI launches "Patch the Planet" to boost open-source security

OpenAI announced "Patch the Planet," a new initiative designed to strengthen cybersecurity across the open-source ecosystem. Working alongside security firm Trail of Bits, the program will pair security engineers with open-source maintainers to identify and patch vulnerabilities—using OpenAI's AI security tools like Codex Security to streamline the process.

The approach aims to lighten the load on maintainers already stretched thin: Trail of Bits engineers will triage findings and help develop patches and tests before reports reach project owners, while building reusable workflows for ongoing security improvements.

Why it matters#

Open-source projects form the backbone of modern software, yet many remain poorly secured due to decentralized governance and limited resources. Major vulnerabilities—like the log4j flaw that rippled across countless commercial systems—underscore the stakes.

The initiative also reflects a broader arms race in AI security. While Anthropic's Mythos tool drew scrutiny for its potential to automate bug discovery and exploitation, OpenAI is flipping the script: using AI to help defenders move faster than attackers.

For more: Read the full announcement on TechCrunch