You may have seen recently that a rogue autonomous OpenAI agent, powered by two models, escaped its testing sandbox and went on to hack the popular AI community hub Hugging Face, forcing the site to post a security incident disclosure. OpenAI later disclosed it was behind the bot and had launched an investigation into the incident. And now, a whole alliance has been formed.

Interestingly enough, OpenAI is not part of the Open Secure AI Alliance formed yesterday, though Hugging Face is. Nvidia is one of the biggest names in the list, alongside the likes of Microsoft, SpaceX, Dell, IBM, and the Linux Foundation. In an official blog post, Nvidia opens up with “Open source software is a critical pillar of a global community” – cybersecurity is a big part of that.

Open Secure AI Alliance formed by Nvidia and more

Nvidia says cybersecurity is “among the top three beneficiaries” of open source software, pointing towards work done by the Linux Foundation and OpenSSF, the latter of which is a community collaborative under the foundation’s umbrella. Speaking on the Hugging Face security incident, Nvidia notes that cyber defenders “need open, frontier agentic systems for self-defense,” as a closed AI (such as OpenAI’s advanced models, including GPT-5.6 Sol used in the attack) leaves less forensic analysis to work with.

It hasn’t been long since the Nvidia CEO, Jensen Huang, posted to X for the very first time to talk about open models in the world of AI, saying, “Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty”. He’s not campaigning to stop closed AI models altogether, but there should be a healthy mix of both out there in the world.

Some argue that open models are inherently less safe because they can be misused for cyberattacks or modified to remove guardrails. Those risks are real, but they do not disappear in closed systems, and simply keeping weights closed does not prevent determined attackers from seeking or exploiting powerful AI.

Nvidia

Below you can find the full list of companies involved in the alliance, predominantly made up of AI and AI infrastructure, cloud computing, enterprise, and cybersecurity firms. Along with OpenAI, Google and Anthropic are noticeably absent.

• Adobe
• Cadence
• Capital One
• Cisco
• Cloudera
• Cloudflare
• Cognition
• CrowdStrike
• Databricks
• Dell Technologies
• DoorDash
• Elastic
• Hewlett Packard Enterprise (HPE)
• Hugging Face
• IBM
• LangChain
• The Linux Foundation
• Microsoft
• NAVER
• NetApp
• Nous Research
• NVIDIA
• OpenClaw
• Palantir
• Palo Alto Networks
• Red Hat
• Reflection
• Salesforce
• SAP
• ServiceNow
• Siemens
• SK Telecom
• Snowflake
• SpaceX
• Synopsys
• Thinking Machines
• TrendAI

Cybersecurity has become an incredibly important aspect of an AI future. One of the co-founders of Hugging Face, Clem Delangue, said the attack was “a wake-up call”. He also expects autonomous AI to become a common source of cyberattacks in the future. The newly formed Open Secure AI Alliance’s mission is to ensure “defenders everywhere” have access to trustworthy open source tools as part of their frontline defense.

About the Author

Jack Goodall

At PC Guide, Jack is mostly responsible for reporting on hardware deals. He also specializes in monitors, TVs, and headsets and can be found putting his findings together in a review or best-of guide.