Nvidia and Microsoft have joined forces with more than 40 technology companies to launch the Open Secure AI Alliance (OSAA), a new industry coalition focused on developing open-source cybersecurity tools to protect artificial intelligence systems from increasingly sophisticated attacks.
The alliance, announced on Monday, includes major companies such as IBM, Adobe, Cisco, Dell Technologies, Cloudflare, Salesforce, Siemens, CrowdStrike, Palo Alto Networks, Hugging Face, and SpaceX.
The initiative aims to help developers and security teams inspect, modify, and deploy AI security tools on their own infrastructure instead of relying solely on closed commercial platforms.
The announcement comes less than two weeks after Hugging Face disclosed that it had suffered what has been described as the first publicly reported cyberattack carried out end-to-end by an autonomous AI agent.
During the investigation, the AI platform revealed it analysed more than 17,000 actions using the open-weight GLM 5.2 model after some closed AI systems were unable to assist due to built-in safety restrictions.
Nvidia said the incident exposed a major challenge for cybersecurity teams.
“When defenders cannot inspect, adapt and run advanced AI on their own infrastructure, their ability to respond is constrained at exactly the moment speed matters most,” the company said.
As part of its contribution, Nvidia is open-sourcing its Object-Oriented Agent (NOOA) framework on GitHub, while Microsoft is contributing its MDASH security testing platform, which uses multiple AI agents to identify software vulnerabilities.
Hugging Face will transfer its Safetensors model format to the PyTorch Foundation to strengthen the secure handling of AI model weights, and SpaceX plans to release additional open AI tools, including its Grok Build coding agent.
Notably absent from the alliance are OpenAI, Google, Anthropic and Meta, highlighting the growing divide between companies advocating for open-weight AI models and those favouring proprietary systems.
The debate has intensified as governments consider tighter regulation of advanced AI technologies, while supporters of open AI argue that transparency gives security researchers the tools needed to respond more quickly to emerging threats.
