Open Secure AI Alliance Proposes SAFE Incident-Reporting Guidelines as Membership Tops 120

The Open Secure AI Alliance, launched by Nvidia, Microsoft, SpaceX, IBM, and dozens of other technology and cybersecurity companies, proposed guidelines called Shared AI Findings Exchange (SAFE) for reporting cybersecurity incidents involving AI agents^2,5. The Linux Foundation published SAFE as a request for comments. The alliance now includes more than 120 organizations. The initiative followed a cyberattack in which Hugging Face reported it was unable to use leading U.S. frontier models for defense due to safety guardrails and instead turned to a self-hosted, open-weight Chinese model^9,10. Google, OpenAI, and Anthropic are absent from the alliance's inaugural membership⁷.