NVIDIA Leads Open Secure AI Alliance to Tackle AI Safety

NVIDIA and a coalition of industry leaders announced the launch of the Open Secure AI Alliance (OSAI) on July 27, 2026. The group aims to advance AI safety and security through open technologies, emphasizing collaboration and transparency in a field increasingly central to global cybersecurity.

Founding members include major players like Adobe, Microsoft, IBM, and Hugging Face, alongside organizations such as the Linux Foundation and OpenSSF. The alliance intends to build open-source tools, models, and frameworks to bolster trust in artificial intelligence. This initiative builds on existing efforts like the Linux Foundation’s Akrites project and OpenSSF’s work to address vulnerabilities in open-source software.

One of the key motivations for the alliance is the belief that open models are essential for effective cyber defense. Open systems allow defenders to inspect, adapt, and deploy AI tools rapidly, a necessity highlighted by recent incidents such as the Hugging Face security breach. In that case, closed AI systems hindered forensic analysis, forcing the company to rely on open models to contain the intrusion.

The Open Secure AI Alliance’s mission is clear: ensure that AI defenses are transparent, adaptable, and broadly accessible. This contrasts with proprietary, closed systems, which can centralize power and create single points of failure. NVIDIA, for example, has contributed its NVIDIA Labs Object-Oriented Agent (NOOA) project, now available on GitHub, to enhance model performance and governance capabilities.

Other contributions include HPE’s zero-trust identity framework SPIFFE/SPIRE, Hugging Face’s Safetensors format for secure AI model storage, and SpaceXAI’s Grok Build terminal-based AI coding agent. These tools exemplify the alliance’s focus on creating a distributed, community-driven defense ecosystem.

Amid this push for openness, the alliance acknowledges the risks associated with open models, such as misuse for cyberattacks. However, its stance aligns with the belief that openness paired with strong safeguards and rapid remediation is the smarter path for defending critical infrastructure.

This launch also comes at a pivotal time in the global debate over AI regulation and national security. As policymakers consider restrictions on open-weight AI models, NVIDIA and its allies argue that such measures could weaken defensive capabilities and increase reliance on a few closed providers. Instead, they advocate for regulatory frameworks that recognize the value of open-source tools in strengthening AI safety.

The Open Secure AI Alliance marks a significant step in addressing the complex challenges of AI cybersecurity. By fostering an open, collaborative approach, the alliance aims to equip defenders with the tools they need to secure AI systems while maintaining transparency and trust.

Image source: Shutterstock