Nvidia's Week-Old AI Security Group Already Shows Progress

The Open Secure AI Alliance (OSAA), an industry group spearheaded by Nvidia and now including over 120 companies, has formed a working group called Shared AI Findings Exchange (SAFE) and is already presenting proposals for public comment, managed by the Linux Foundation. The proposals cover confidential reporting of AI security incidents, alerting affected parties, and blame-free analysis. Members are also contributing open-source technology, such as Nvidia's Garak LLM vulnerability scanner, Okta's agent identity tech, Red Hat's agent governance, and Amazon's Strands Agents and Cedar authorization language. The group includes big names like Adobe, BlackRock, Cisco, Intel, Microsoft, and Visa, but notable absences are Anthropic, OpenAI, and Google, although OpenAI and Google signed the original open letter supporting open source AI. The group operates at AI speeds, forming quickly in response to potential regulatory threats, and aims to enhance AI safety through openness.