Just one week after the establishment of the Open Secure AI Alliance (OSAA) spearheaded by Nvidia, the industry group has grown to over 120 companies and is already delivering its first concrete proposals. Operating at high speed during the Black Hat conference in Las Vegas, the coalition launched a working group called SAFE, managed by the Linux Foundation, to establish collaborative safety standards.
The initial proposals focus on confidential reporting of AI cybersecurity incidents and blame-free post-mortems to foster collective learning. At the same time, major tech players including Amazon, Red Hat, Okta, and Nvidia are contributing open-source tools and technologies to help enterprises secure their AI agents and defend against potential vulnerabilities or attacks.
Despite broad participation from giants like Microsoft, Intel, and Cisco, notable absences include Google, OpenAI, and Anthropic, even though Google and OpenAI supported the initial open letter backing open-source AI. The rapid progress of the alliance marks a significant step in strengthening the U.S. open AI ecosystem and establishing robust industry-wide safeguards.
- OSAA grew to over 120 member companies within its first week.
- The new SAFE working group addresses AI cybersecurity incident reporting.
- Major firms like Amazon and Nvidia are actively contributing open-source security tools.
- Google, OpenAI, and Anthropic are currently notable absences from the alliance.
Sources:
