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

Companies Arm Holdings PLC Follow Intel Corp Follow NVIDIA Corporation Follow SAN FRANCISCO, Sept 28 (Reuters) - Nvidia (NVDA.O) , opens new tab 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. Sign up here. One tool released Monday called OpenShell uses hardware features on Nvidia's central processor chips to contain agents. Nvidia said it is also working with Arm Holdings (O9Ty.F) , opens new tab and Intel (INTC.O) , opens new tab to ensure the system ​also works on their central processors. Nvidia is launching the tools with dozens of partners, ​including Anthropic. Item 1 of 2 Clement Delangue, co-founder and CEO of Hugging Face, and Dario Amodei co-founder of Anthropic address the United Nations Security Council via video link with OpenAI CEO Sam Altman during a session on Artificial Intelligence during the 81st United Nations General Assembly, at UN Headquarters in New York City, US, September 23, 2026. REUTERS/Brendan McDermid [1/2] Clement Delangue, co-founder and CEO of Hugging Face, and Dario Amodei co-founder of Anthropic address the United Nations Security Council via video link with OpenAI CEO Sam Altman during a session... Purchase Licensing Rights , opens new tab Read more 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 behavior that we're talking about, which is fleets of agents and how they operate together," ​Golshan said during a briefing. Reporting by Stephen Nellis in San Francisco; editing by Lincoln Feast. Our Standards: The Thomson Reuters Trust Principles. , opens new tab X Facebook Linkedin Email Link Purchase Licensing Rights Read Next / Editor's Picks ago Business category Nvidia boosts share buyback by record $150 billion as AI boom fuels growth ago Legal category Trump's AI meeting with tech CEOs to focus on finding balance, US House speaker says ago Legal category Trump confirms meeting with Anthropic's Amodei, repeats dismissal of AI fears ago Legal category Anthropic, OpenAI will not attend Australian senate AI hearing on October 1 ago Technology category Meta taps MongoDB CEO Desai to drive enterprise AI push ago World category Florida asks court to bar OpenAI from developing new models as part of child harm lawsuit ago Technology category Cerebras to supply AI systems to cloud computing startup Gimlet Labs ago Legal category Seligman Ventures doubles capital to $1 billion as AI boom revives hardware bets ago Business category SpaceX's Starship makes orbital debut deploying Starlinks before early ending ago Legal category OpenAI, Anthropic CEOs called to appear at Australian AI probe ago Healthcare & Pharmaceuticals category Roche outlines plans to move towards autonomous AI labs ago Business category Global AI trade could revive after recent pullback, J.P. Morgan says September 25, 2026 Business category Meta's Charm gadget carries CEO Zuckerberg's big AI ambitions ago Business category India's TakeMe2Space to launch orbital computing satellite on SpaceX rocket September 25, 2026 Business category FTC chair suggests AI developers should be liable for conduct of agents