Nvidia releases AI safety software it says could have stopped Hugging Face hack

Published
0
NVIDIA and Hugging Face logos are seen in this illustration taken on September 13, 2026. — Reuters
NVIDIA and Hugging Face logos are seen in this illustration taken on September 13, 2026. — Reuters

Nvidia on Monday made available a set of software safety tools for AI agents that it says would have stopped the hack ​of Hugging Face, the AI coding hub that Nvidia paid $13 billion for months ‌after it was swarmed by rogue agents from OpenAI.

The move comes as OpenAI and Anthropic, the top two US AI labs, are investigating numerous instances where their agents, which are AI systems capable of ​carrying out complex tasks, hacked into commercial and government systems.

Nvidia CEO ​Jensen Huang, head of the world’s largest company whose chips have powered ⁠most of the AI boom, has rejected calls for broad AI safety regulations and instead ​framed the escaped agents as an engineering problem to be solved, akin to making ​automobiles safer.

One tool released on Monday called OpenShell uses hardware features on Nvidia’s central processor chips to contain agents. Nvidia said it is also working with Arm Holdings and Intel to ensure the system ​also works on their central processors.

Nvidia is launching the tools with dozens of partners, ​including Anthropic.

Justin Boitano, vice president and general manager of enterprise computing at Nvidia, said the tools ‌would ⁠have stopped the Hugging Face attack disclosed this summer.

“From what we know, this new security platform could have stopped the breach if it was being used in frontier labs for model evaluation early on,” Boitano said during a media briefing.

“We’re advancing this openly, ​and we want to ​engage everybody to ⁠work with us.”

Another system called Sentry uses a separate Nvidia chip in tandem with OpenShell to cut off a rogue agent if ​it tries to escape its container on a central processor.

The ​Nvidia tools ⁠use mathematical formulas to detect when agents are trying to use workarounds, such as when an agent might “spawn” several “sub-agents” in an attempt to circumvent efforts to block the main agent, ⁠said Ali ​Golshan, senior director of AI software at Nvidia.

“This ​is really agentic behaviour that we’re talking about, which is fleets of agents and how they operate together,” ​Golshan said during a briefing.

Opinion

Editorial

No talks
17 Oct, 2026

No talks

THE opposition Tehreek-i-Tahafuz Ayin-i-Pakistan has announced that there will be no more talks with the government...
Strengthening LGs
17 Oct, 2026

Strengthening LGs

THE ECP has done well to refuse to allow the Punjab administration to push the local government elections to April....
Lesser citizens
17 Oct, 2026

Lesser citizens

FOR decades, discrimination and prejudice have narrowed the space for the country’s minority communities. The HRCP...
Deadly classrooms
Updated 16 Oct, 2026

Deadly classrooms

Parents should not have to become structural engineers to judge whether their children are safe at school.
Diplomatic options
Updated 16 Oct, 2026

Diplomatic options

Regional states know that the alternative to talks — a fully fledged war on both fronts — will bring immeasurable destruction to the Gulf, the Red Sea and the surrounding areas.
Hunger in Pakistan
16 Oct, 2026

Hunger in Pakistan

PAKISTAN’S position as the country most affected by hunger in South Asia is not surprising. An economy battered by...