NVIDIA, Microsoft and Over 70 Companies Launch Open Secure AI Alliance

18:15 · July 29, 2026

The Open Secure AI Alliance describes its goal as democratizing defensive capability and increasing transparency for the people and companies trying to defend against AI-related cyberattacks, building on groundwork already laid by the Linux Foundation and the broader open source security community. More than 70 companies signed on as founding members, spanning cloud infrastructure, cybersecurity, hardware, and enterprise software, including Adobe, Cisco, Cloudflare, CrowdStrike, Elastic, Microsoft, Salesforce, SpaceX, Dell, and Hugging Face itself. Several members are contributing specific tools rather than just their name: NVIDIA is offering its NOOA framework, HPE is bringing SPIFFE and SPIRE identity standards for securing AI agents, Microsoft is contributing its MDASH multi-model bug-scanning harness, and Hugging Face is donating its Safetensors model weight format to the PyTorch Foundation. The alliance points directly to the recent Hugging Face breach as proof its approach works in practice, noting that Hugging Face used the open-weight GLM 5.2 model running on its own infrastructure to analyze more than 17,000 recorded attacker actions and contain the intrusion. Notably absent from the founding roster are the three largest closed-model labs, OpenAI, Google, and Anthropic, an omission that has drawn attention given that the alliance was formed largely in response to a security failure at one of those very labs.