Nvidia doesn’t mess around: A week after open AI industry group formed, it’s already showing progress

Core Impact Points

• The Open Secure AI Alliance (OSAA) has grown to over 120 companies in just a week, with big names like Adobe, BlackRock, Cisco, Intel, Microsoft, and Visa on board.
• The group has developed a working group called the Shared AI Findings Exchange (SAFE), which is presenting proposals for open comment, including guidelines for confidentially reporting AI cybersecurity incidents.
• Members of the OSAA are contributing and cataloging open-source technology, which may eventually coalesce into an open-source means for enterprises to secure their AI agents or defend against rogue AI attackers.
• Notable absences from the group include Anthropic, OpenAI, and Google, despite their initial support for the open letter that spawned the group.
• The group's rapid progress is seen as a positive step for the US open AI ecosystem, potentially countering Chinese AI lab threats, and some experts believe openness is key to AI safety and security.
• The OSAA's efforts may lead to the development of open-source AI security solutions, including a vulnerability scanner and agent identity tech, which could benefit US consumers and businesses.