Nvidia and a number of tech giants on Monday launched a new artificial intelligence security initiative focused on open models, as the fallout from a cyberattack by fraudulent OpenAI models continues.
Last week it emerged that the target of the attack, startup Hugging Face, was unable to use leading US border models for self-defense because the guardrails did not differentiate between attacker and defender. Instead, they turned to a self-hosted, open Chinese model that wasn’t bound by the same restrictions.
As U.S. lawmakers increasingly consider how to curb the increasing adoption of Chinese AI models, the most advanced of which are open-weight models, tech giants have launched an initiative aimed at developing and sharing open AI tools.
Open models can be downloaded, modified and self-hosted, unlike closed models – including border systems from Anthropic and OpenAI – which can only be accessed through specific infrastructure.
“The Open Secure AI Alliance will work to remediate and disclose vulnerabilities using open technologies,” Nvidia said in a statement. “The recent Hugging Face security incident made it clear: cyber defenders need open, cross-border agent systems for self-defense.”
The other members of the alliance include Nvidia Microsoft, SpaceX, Palantirand dozens of other technology companies from the US and Europe.
Table of Contents
The push to curb Chinese AI
There are growing calls for action to restrict access to models from Chinese AI companies, which are accused of campaigns to extract information from U.S. rivals’ systems, known as “distillation” – when a model extracts knowledge from a better-trained model.
Last week, Treasury Secretary Scott Bessent threatened sanctions against Chinese companies that carry out distillation attacks against U.S. companies.
“There is a real possibility that the U.S. government will impose restrictions on the Chinese models,” Chris McGuire, senior fellow for China and emerging technologies at the think tank Council on Foreign Relations, told CNBC.
That could include a ban on transactions involving the models, such as buying tokens through an API or U.S. companies hosting the model in the cloud and charging customers for inference, he said.
“In Washington, this is not a debate about open source vs. closed source, but a debate about whether or not to tolerate Chinese IP theft,” McGuire said. “Any measures would focus on Chinese companies, not the open source ecosystem.”
However, since most of the most powerful open source models are developed by Chinese companies, there are concerns about limitations.
Last week Nvidia, Microsoft, MetaPalantir and more than 20 other companies released a letter urging policymakers to avoid “premature restrictions” on open AI models that would “stifle competition or spur innovation abroad.”
The OpenAI hugging face incident demonstrated a “practical truth,” Nvidia said. “If defenders cannot inspect, adapt, and operate advanced AI on their own infrastructure, their ability to respond is limited at the very moment when speed matters most.”
Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.
https://www.cnbc.com/2026/07/27/nvidia-ai-initiative-openai-cyber-attack.html
