Logos of founding members of the Open Secure AI Alliance, including NVIDIA, Microsoft, and IBM. — Leading technology companies unite to establish the Open Secure AI Alliance for standardized AI security.

Industry Leaders Unite in Open Secure AI Alliance to Strengthen Global AI Safety and Security

NVIDIA and a coalition of global technology leaders have launched the Open Secure AI Alliance to develop and share open-source defensive tools and research. This movement aims to provide defenders with the transparency, adaptability, and sovereign control necessary to safeguard critical infrastructure against AI-enabled threats.

RMN Digital AI Desk
New Delhi | July 27, 2026

In a landmark move for the future of artificial intelligence, NVIDIA and dozens of inaugural partners have established the Open Secure AI Alliance. This coalition brings together leaders from cloud computing, cybersecurity, and AI research to build a shared foundation of open tools designed to promote responsible AI use and global trust.

The alliance includes industry giants such as Adobe, Cisco, Dell Technologies, IBM, Microsoft, and Salesforce, alongside open-source foundations like the Linux Foundation. By building on existing initiatives like the Akrites project and the OpenSSF community, the alliance seeks to remediate and disclose vulnerabilities using open technologies.

The Open Secure AI Alliance democratizes defense, providing the transparency and localized control needed to prevent single points of failure in global AI infrastructure.

The Necessity of Open Defense

A core philosophy of the alliance is that open models and harnesses are essential for robust cybersecurity. While closed systems have their place, open-source tools democratize defensive capabilities and ensure there is no single point of failure in critical infrastructure.

The practical necessity of this approach was recently highlighted by a security incident at Hugging Face. When closed AI tools—unable to distinguish between attackers and defenders—blocked essential forensic analysis, defenders successfully utilized an open-weight model on their own infrastructure to contain the intrusion. The Open Secure AI Alliance aims to ensure that such frontier defensive tools are always available for defenders to inspect, adapt, and run locally.

Through open-source projects like NOOA and MDASH, the alliance is building a complete defense stack to safeguard the future of agentic AI.

Collaborative Innovation and Research

The alliance is already contributing significant resources to the open-source community:

NVIDIA has released the NVIDIA Labs Object-Oriented Agent (NOOA) project on GitHub to make advanced AI safety more accessible for agent harnesses.
Microsoft contributed MDASH, a multi-model scanning harness that orchestrates AI agents to discover exploitable bugs.
HPE is advancing the SPIFFE/SPIRE zero-trust identity framework to verify AI agents cryptographically.
IBM and Red Hat are extending security across the supply chain through the Lightwell project.
SpaceXAI has open-sourced the Grok Build coding agent and plans to open-source model weights to support the broader research community.

A Call for Open Standards

The alliance also serves as a collective voice for policymakers and regulators, urging them to recognize open frontier AI systems as defensive assets rather than liabilities. The coalition argues that blanket restrictions on open systems could weaken global defensive capacity and risk concentrating power within a few opaque providers. By investing in shared open infrastructure—such as datasets and red-teaming tools—the alliance believes the age of AI can be one of resilience and shared security.