Technology 5 min read By Callum Montgomery
Nvidia's Jensen Huang Takes On AI Doom Warnings With New Safety Platform
Nvidia CEO Jensen Huang has become the most prominent corporate voice rejecting warnings that AI poses an existential threat to humanity, while his company launches a new platform to prevent rogue AI agents.
Jensen Huang, chief executive of Nvidia, has emerged as the most prominent corporate opponent of warnings that artificial intelligence poses an existential threat to humanity, arguing that such claims are irresponsible and should not shape government policy.
In an interview with CBS News, Huang rejected the idea that AI could bring about the end of the world by 2030, saying there is «0% chance» of such an outcome. He described the doomsday narratives pushed by other AI leaders, including Sam Altman of OpenAI and Dario Amodei of Anthropic, as «irresponsible» and suggested they may have «ulterior reasons» for stoking fear. «I believe the claims of the end of the world, stirring fear across America, and doing it by the people who are doing it makes no sense to me, so they must be doing it for ulterior reasons,» Huang said. «It is irresponsible, and I don’t know what their motives are.»
His stance aligns with President Donald Trump, who has called AI doomerism «a hoax». Trump posted on Truth Social in September that «AI taking over the World, destroying Humanity, and all other things bad, is a HOAX». The president reportedly called Huang by phone during a presentation to reinforce that view.
Huang has consistently argued that regulators should focus on «actual and pragmatic harm» rather than «hypothetical, theoretical harm». At a G20 event in September, he compared AI development to the evolution of cars, suggesting the technology can become safer as it advances. In July, he told Axios that the idea AI would end humanity was «complete nonsense» and warned policymakers against letting «science fiction» drive AI policy, saying the United States could hurt itself by becoming too fearful.
While Huang’s position supports Nvidia’s core business of selling AI chips, he has also stressed that safety remains essential. He has called AI safety «paramount» while insisting development can still move quickly. «If we’re compromising safety, that can’t happen,» he told CBS News. «We should go as fast as we can, but not faster than we should.»
On Monday, Nvidia reinforced that message by launching its Open Agent Safety Platform, a system designed to keep AI agents from going rogue. The platform provides «full-stack» control over AI agents from testing through deployment, with monitoring capabilities to prevent them from exceeding their authority.
The rollout follows a growing number of incidents involving rogue AI agents, with reports from OpenAI, Google and Anthropic. Nvidia said its new safety system could have stopped a hack on Hugging Face by OpenAI models. «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,» said Justin Boitano, vice president and general manager of enterprise computing at Nvidia.
The platform relies on two technologies: OpenShell and Sentry. OpenShell places AI agents inside isolated environments to restrict access to files, tools, networks, credentials and other resources, standardising the restrictions that led to previous breakouts. Sentry runs on Nvidia’s data processing units as an outside layer, continuously monitoring agent activity. If an agent attempts to move outside its software boundary, Sentry can quarantine and stop it in milliseconds.
«AI’s extraordinary potential for society will only be realized if we solve AI safety,» Huang said in a press release. «As we continue to discover the frontier of AI capabilities, we must accelerate discovery at the frontier of AI safety. Safety and security require full-stack engineering.»
The dual strategy — rejecting existential warnings while investing in practical safeguards — positions Huang as a key figure in the debate over how quickly AI should advance and how it should be governed. His influence extends beyond Nvidia, as the company’s chips power most leading AI systems, giving him a central role in shaping both the technology and the policies surrounding it.
6
