OSAA Launches SAFE Group With 120-Plus Members as US-China AI Ban Fears Spur Open Security Push
Updated
Updated · TechCrunch · Aug 4
OSAA Launches SAFE Group With 120-Plus Members as US-China AI Ban Fears Spur Open Security Push
1 articles · Updated · TechCrunch · Aug 4
Summary
The week-old Open Secure AI Alliance has set up SAFE—the Shared AI Findings Exchange—and is already circulating draft proposals for public comment at the Black Hat conference in Las Vegas.
The proposals focus on practical AI security basics: confidential incident reporting, alerts to affected users, and blame-free postmortems, with the Linux Foundation managing the process.
More than 120 companies have joined OSAA, and members are cataloging open-source security tools from Nvidia, Okta, Red Hat and Amazon that could evolve into shared defenses for AI agents and enterprise systems.
OpenAI, Google and Anthropic are notable nonmembers, even though OpenAI and Google signed the open letter that helped spark the alliance.
That letter—backed by more than 200 tech companies and championed by Nvidia—followed concern that the Trump administration might ban Chinese open-weight AI models, pushing the industry to organize quickly around open AI security.