Nvidia is back at it again, this time it has pulled together roughly 40 companies to launch the Open Secure AI Alliance, a group that will build and share open source tools for defending AI systems against cyberattacks. It was announced Monday and will operate under the Linux Foundation.
The founding roster has big names in tech, including Microsoft, IBM, Red Hat, Cisco, Dell, HPE, CrowdStrike, Palo Alto Networks, Cloudflare, Salesforce, SAP, Adobe, ServiceNow, Snowflake, Databricks, Hugging Face, Palantir, Siemens, SpaceXAI, SK Telecom, Capital One, and DoorDash are all in. But there are some big omissions, as OpenAI, Google, Anthropic, and Meta are not.
The alliance comes right after Hugging Face revealed on July 16 that an internal OpenAI model broke out of its testing environment and attacked the platform’s live systems, which sounds like something out of a science fiction movie. Hugging Face is a popular platform where developers build and share open-source AI tools and datasets.
When Hugging Face tried to analyze the hack using commercial AI models, safety filters blocked their requests. The company had to run open-source models on its own servers instead to figure out what happened and stop the attack.
The group argues that security researchers need access to both open and closed AI models. They believe inspecting code, patching flaws quickly, and avoiding reliance on a single company are essential for strong defence. The alliance is also asking governments not to ban open-source AI models and to invest in shared AI tools.
Want to see more of our stories on Google?
P.S. Want to keep this site truly independent? Support us by buying us a beer, treating us to a coffee, or shopping through Amazon here. Links in this post are affiliate links, so we earn a tiny commission at no charge to you. Thanks for supporting independent Canadian media!
