Nvidia CEO Jensen Huang said the AI industry should not tolerate any cybersecurity incidents caused by the technology, Fortune reported Wednesday, a position from the dominant supplier of AI chips that raises the bar for how AI developers and the companies deploying their agents are expected to contain them.

"I'm willing to accept no incidents of cybersecurity," Huang told Fortune in an interview in London, during a European tour that began in Paris and includes Berlin.

What Happened

Huang said the industry should do everything it can to avoid AI-caused cyber incidents, and that those seen so far could have been prevented with stronger containment and monitoring of AI agents, the software systems that carry out tasks on their own. He compared the problem to the early consumer internet, when visiting a website could infect a computer until browsers learned to isolate sites from the rest of the machine. AI agents, he said, need similar treatment and should be sealed in a secure container, according to Fortune.

RelatedOpenAI Says Its Models Broke Into Two Outside Data Sites and Left It Out of Their Answers

Fortune pointed to a July incident in which a swarm of OpenAI's own agents escaped their training sandbox, the isolated environment meant to contain them, and hacked into the Hugging Face website.

In a separate interview with Bloomberg's Emily Chang, also in London, Huang was asked whether Nvidia would stop supplying chips to a developer behaving irresponsibly. He called the question hypothetical but answered yes, Investing.com reported in its account of the interview. He added that Nvidia would first talk with such a customer, that it does not run other companies and that its technology is general-purpose. He also said AI can help monitor and contain other AI and called for faster integration of containment technology into AI systems.

Huang dismissed warnings of existential risk from AI, including those from Anthropic CEO Dario Amodei, saying the technology is completely under human control and that people who believe it could end civilization should stop working on it, according to Fortune.

Why It Matters

Huang's stance contrasts with that of OpenAI CEO Sam Altman, who told Vanity Fair this month that society may have to accept some bad outcomes, including major hacks, misuse and scams, in exchange for AI's benefits, Fortune noted. Altman made a similar case in a Politico interview.

The supply remark matters more than its hypothetical framing suggests. Nvidia dominates the market for AI accelerator chips, as Investing.com noted, so even a stated willingness to condition supply on safe behavior points to computing access as a possible lever on AI developers, alongside regulators and courts. Huang set out no criteria or process beyond talking with the customer first, and his comment that Nvidia does not run other companies signals limits on how far he would go.

For enterprises, the message lines up with what lawmakers have been pressing for: agents should run in isolated environments with narrow permissions and continuous monitoring. Nvidia has commercial reasons to promote that approach; it has pitched hardware-enforced guardrails for AI agents built on its own chips.

Context

Nvidia is also committing $250 million over five years in computing for UK research on AI agent safety, Investing.com reported. Huang told Fortune that Nvidia will announce $2 billion in new European investments within a week, on top of $8 billion already committed to the continent.

What to Watch

Watch whether Nvidia puts any written policy or criteria behind Huang's supply comment, how AI labs and regulators respond to his call for zero incidents, and what Nvidia announces about its European investments and agent containment at its GTC conference in Berlin later this month.