Linux Foundation, Tech Giants Launch Akrites to Defend Open Source Against AI-Powered Attacks
The Linux Foundation launched Akrites with 19 founding organizations to speed up fixing critical open-source vulnerabilities before AI-powered attackers can exploit them. Frontier AI models can now find confirmed flaws in minutes, compressing the window between discovery and exploitation to near real time. Akrites replaces the old noisy, slow coordinated-disclosure model with one confidential Security Incident Response Team that routes validated findings upstream in a predictable way, returns fixes to original repos, and can act as maintainer of last resort for abandoned packages. Leaders said success means not just publishing patches, but getting them deployed before exploits emerge. OpenAI also recently launched a related effort, Patch the Planet, focused on AI-assisted discovery and patching, while Akrites provides the broader coordination layer. Alpha-Omega is seeding funding for the program.
