In response to security challenges highlighted during the recent Hugging Face incident, where closed frontier AI models obstructed essential forensic analysis, a new initiative called the Open Secure AI Alliance has been formed. This alliance aims to foster collaboration among industry leaders, including NVIDIA, to promote open innovation and shared tools for AI safety and security. As emphasized by industry experts, attackers utilize advanced frontier AI, necessitating the development of a comprehensive ecosystem of both open and closed AI models to effectively defend against such threats.

Jensen Huang: Jensen Huang is the founder and CEO of NVIDIA. In the context of this news, he publicly advocated for a frontier AI ecosystem combining open and closed models, citing the Hugging Face incident where an open-weight model aided containment after closed AI blocked forensics, and announced the creation of the Open Secure AI Alliance to address this gap.
Open Secure AI Alliance: The Open Secure AI Alliance is a new industry coalition launched by NVIDIA and dozens of partners including Microsoft, Hugging Face, CrowdStrike, IBM, and the Linux Foundation to develop and share open AI security tools, models, and research. It builds on existing open-source security efforts like the Linux Foundation’s Akrites initiative to strengthen defenses against AI-enabled threats. The alliance was formed to ensure defenders have access to frontier open models that support forensics and containment in incidents where closed systems fall short.

Incident Response: During the recent Hugging Face security incident, closed frontier AI models blocked essential forensic analysis while an open-weight frontier model successfully helped contain the intrusion.
Alliance Formation: The Open Secure AI Alliance was established to promote open innovation, shared tools, and a distributed community of defenders for AI safety and security.
Industry Collaboration: NVIDIA and numerous technology leaders across cloud, cybersecurity, and AI research have joined as inaugural partners to advance open technologies for safeguarding AI systems and agents.