The Open Secure AI Alliance (OSAA), launched by Nvidia just a week ago, is already turning momentum into action. The group has expanded to more than 120 companies and has introduced a new working group called the Shared AI Findings Exchange (SAFE).
Managed by The Linux Foundation, the initiative is now circulating proposals for public feedback. Its early focus is practical: creating clear ways to report AI security incidents, notify affected parties, and review events in a blame-free way that helps the wider ecosystem learn faster.
Members are also pooling open-source tools and technical components that can strengthen AI safety. Nvidia has highlighted its open models and the Garak vulnerability scanner, while other participants are contributing work on agent identity, governance, authorization, and secure agent development.
The alliance includes major names such as Adobe, BlackRock, Cisco, Intel, Microsoft, and Visa. At the same time, the group's formation reflects a broader industry push for open and secure AI development, especially as open-weight models continue to gain attention across the sector.
With proposals already in motion and collaboration growing quickly, OSAA is positioning itself as a practical framework for safer enterprise AI. If this pace continues, it could help shape a more transparent and resilient future for open AI systems.
