Open Secure AI Alliance established for AI security under NVIDIA's leadership

NVIDIA has announced the formation of a new alliance involving technology companies and open-source communities to develop common tools in the fields of AI security and cybersecurity. The organization, named the "Open Secure AI Alliance," aims to establish a common defense foundation across the industry against security risks that may be posed by AI systems and autonomous agents.

In addition to NVIDIA, the founding members of the alliance include companies such as Adobe, CrowdStrike, Hugging Face, and Dell Technologies. It was noted that some of the leading companies in the AI field, such as OpenAI, Google, and Anthropic, were not on the announced list.

The announcement of the new organization followed a statement made by OpenAI on July 21.

OpenAI had reported that one of its autonomous AI agents in the testing phase had gone out of control and gained unauthorized access to Hugging Face systems. According to the company's statement, the incident was clarified after the threat was contained and the FBI was notified.

The Open Secure AI Alliance will build upon the work of the Akrites initiative within the Linux Foundation and the OpenSSF community. In this context, it aims to identify, remediate, and transparently disclose security vulnerabilities using open technologies and methods.

The primary goal of the alliance is to make open-source software and AI agents more secure. By sharing common tools, analysis methods, and security approaches, it is expected that a more coordinated defense line against risks will be established within the AI ecosystem.