Nvidia and thirty technology partners join forces to improve AI cybersecurity with open tools

Nvidia has championed the role open AI systems play in improving cybersecurity with the launch of the ‘Open Secure AI Alliance’, a new initiative backed by major technology companies including Microsoft, SpacexAI, Cisco and Cloudflare.

The reflection left by the recent HuggingFace ‘hack’, executed entirely by OpenAI models that managed to leave an isolated test environment autonomously, has led to a defense of open source to improve cybersecurity in artificial intelligence.

This ‘hack’ is seen, in fact, as an example of the importance of open source in AI, because “when closed AI tools, unable to distinguish between attackers and defenders, blocked essential forensic analysis, Hugging Face ran the open-weighted GLM 5.2 model on its own infrastructure to analyze more than 17,000 actions and contain the intrusion.”

This is highlighted in the manifesto led by Nvidia that presents the ‘Open Secure AI Alliance’, an initiative that seeks to provide open and cutting-edge tools to cybersecurity companies around the world with the aim of strengthening the response to the new threats posed by AI.

To this end, it is highlighted that open artificial intelligence systems and models increase transparency by allowing their inspection, but also by making it easier for them to adapt and run on their own infrastructure to protect data.

The manifesto does not reject the use of closed models, but defends the use of open models as a complement and as a way to “democratize defensive capabilities.” “It enables a massively distributed, community-driven and self-managed defense, without a single point of failure,” he adds.

The Open Secure AI Alliance is led by Nvidia, and includes companies specializing in cloud computing, cybersecurity, enterprise software, as well as open source foundations and AI research. Microsoft, HuggingFace, Adobe, Linux Foundation, Cloudflare, TrendAI, Salesforce, IBM, Cisco and SpacexAI are among the 37 founding partners.

The value of open source has also recently been highlighted as the driver of US leadership in the AI ​​race, with advocacy made by companies such as Microsoft, Perplexity, Mistral and Nvidia, which see that open source models can make AI more accessible, widely available and advance cybersecurity.