Open Secure AI Alliance Triples Membership as Industry Unites on AI Security

The Open Secure AI Alliance (OSAA) has rapidly expanded its membership, tripling in size just days after its launch as technology companies rally around the need for stronger AI security. Led by NVIDIA, the alliance now includes more than 100 organizations spanning cloud computing, cybersecurity, enterprise software, research institutions, and open-source communities. The coalition aims to develop open standards, tools, and frameworks that help organizations build, deploy, and secure increasingly powerful AI systems and autonomous AI agents.

The alliance was formed in response to growing concerns that AI-powered cyberattacks are becoming more sophisticated and that defenders need transparent, collaborative security tools to keep pace. Rather than relying solely on proprietary technologies, OSAA promotes an open approach in which researchers and companies can inspect, improve, and share defensive AI technologies. Founding and new members include major names such as Microsoft, IBM, Cisco, Cloudflare, CrowdStrike, Red Hat, Hugging Face, Palo Alto Networks, and the Linux Foundation, creating one of the largest collaborative AI security initiatives to date.

The alliance’s work covers every layer of AI security, from protecting AI models and agent identities to securing software supply chains, evaluating AI behavior, scanning for vulnerabilities, and developing safer deployment frameworks. Among its early technical contributions are NVIDIA’s NOOA (NVIDIA Labs Object-Oriented Agents) framework, Microsoft’s MDASH vulnerability-scanning harness, IBM and Red Hat’s Lightwell software integrity project, and Hugging Face’s Safetensors initiative for securely distributing AI model weights. Together, these projects are intended to create an open ecosystem that organizations can adapt to their own security requirements.

One notable aspect of the alliance is the absence of several leading frontier AI developers, including OpenAI, Google, and Anthropic. While the coalition argues that open AI technologies are essential for effective cyber defense, it also states that open and closed AI models should complement each other rather than compete. The alliance believes organizations should have the flexibility to choose the best combination of AI systems while avoiding dependence on a single vendor or proprietary platform.

The rapid growth of the Open Secure AI Alliance highlights the increasing importance of cybersecurity as AI becomes deeply integrated into businesses, governments, and critical infrastructure. As AI agents gain greater autonomy and attackers adopt more advanced techniques, collaboration between technology companies is expected to play a crucial role in protecting digital systems. By expanding its membership and encouraging open innovation, the alliance aims to accelerate the development of secure, transparent, and trustworthy AI technologies that can support the next generation of intelligent applications.