Nvidia has partnered with 36 technology companies to launch a new industry collective focused on developing open-source artificial intelligence tools for cybersecurity, a move that notably excludes some of the sector's most prominent AI developers. The alliance, announced this week, brings together major technology firms including Microsoft and IBM to create standardized defense frameworks and security technologies that organizations can deploy independently on their own infrastructure.
According to Decrypt, more than 40 technology companies and organizations have joined the initiative, dubbed the Open Secure AI Alliance. However, CoinDesk reports the group comprises 37 members including the chipmaker itself, noting that major AI players OpenAI, Anthropic, and Google are absent from the participant list.
The coalition aims to address growing concerns about AI-driven cyber threats by providing defenders with open-source tools and standards they can operate themselves, rather than relying on proprietary systems controlled by third parties. As reported by CoinDesk, the alliance emphasizes that defenders need AI systems they can run independently, reflecting a strategic priority for operational autonomy in security operations and reducing reliance on external cloud providers for sensitive defensive capabilities.
The exclusion of OpenAI, Google, and Anthropic marks a significant divergence in how major technology firms are approaching AI security collaboration. While these companies have developed some of the most advanced large language models and AI systems currently available, the new alliance appears focused on enabling organizations to maintain direct control over their defensive AI capabilities through self-hosted, auditable solutions rather than cloud-dependent services that may present data exposure risks.
Microsoft and IBM's participation alongside Nvidia signals substantial enterprise backing for the open-source security initiative. The alliance's formation comes as businesses and government agencies grapple with securing AI systems against adversarial attacks while maintaining strict data sovereignty over sensitive cybersecurity operations and classified threat intelligence.
By prioritizing open-source development, the group aims to create interoperable security tools that can function across different platforms and environments without vendor lock-in. This approach addresses supply chain security concerns and allows organizations to inspect, modify, and harden their defensive AI tools according to specific operational requirements. The initiative represents a distinct philosophical split in the AI industry between closed, proprietary security models and transparent, community-driven defense mechanisms that prioritize user control and transparency over the convenience of managed services.