Nvidia CEO Jensen Huang has rejected claims that chip export restrictions will significantly slow rival AI programs, arguing that countries such as China already have the resources to build advanced systems. He linked this to developments such as Anthropic’s “Claude Mythos,” which he said highlight the need for the US and China to work together on AI so researchers can agree on how to safely use increasingly powerful technology.
“We want the United States to win. But I think having a dialogue and having a research dialogue is probably the safest thing to do,” he said, speaking on the Dwarkesh Patel podcast.
According to Huang, AI growth will not only depend on advanced chips but also on a broader set of factors including algorithms, energy, and talent, which are just as necessary as hardware for advancing artificial intelligence. He pushed back against the idea that restricting chip exports can meaningfully constrain AI development.
Also Read: Washington Calls It “Promoting Secure Exports.” But Is It a Chokehold on Global AI?
He pointed out that even older-generation chips, such as 7nm technology, remain highly capable. He argued that scale and system design can help offset gaps in semiconductor technology.
The Nvidia CEO also highlighted the importance of software and algorithms over raw hardware gains. While Moore’s Law delivers incremental improvements, breakthroughs like mixture-of-experts architectures and advanced attention mechanisms can drive significant performance gains, he said.
His remarks come amid growing attention around Anthropic’s Claude Mythos, which is claimed to identify software vulnerabilities. However, these claims remain unverified, with experts divided over whether such systems represent a genuine cybersecurity breakthrough or an overstated risk. “The amount of capacity and the type of compute (Mythos) was trained on is abundantly available in China, so you just have to first realize that chips exist in China.”
“They (China) have datacenters that are sitting completely empty, fully powered,” Huang added. “You know, they have ghost cities, they have ghost data centers too. They have so much infrastructure capacity. If they wanted to, they [could] just gang up more chips…Their capacity of building chips is one of the largest in the world.”
At the same time, Huang cautioned that placing restrictions on U.S. technology could result in a counterproductive outcome, warning that limiting exports would push many worldwide developers toward alternative ecosystems that would diminish the influence of the United States over the evolution of AI standards.
“This is an area that is glaringly missing because of our current attitude about China as an adversary,” he said. “It is essential that our AI researchers and their AI researchers are actually talking.”
Huang reiterated that it is critical for the U.S. to maintain a leading position in artificial intelligence by expanding compute and innovation; however, this should not be done by disenfranchising competitors.
He proposed a compromise, encompassing both leading and global participation, which will ensure the dominance of the American AI stack as the AI technology develops across the globe.
“Victimizing them, turning them into an enemy, likely isn’t the best answer,” he added. “It is essential that we try to both agree on what not to use the AI for.”
More broadly, Huang’s comments suggest that developments like Mythos highlight that AI progress cannot be contained through hardware restrictions alone, making it increasingly important for the U.S. and China to work together on AI.
Huang’s remarks contribute to the ongoing debate regarding whether the United States should seek its position of leadership with artificial intelligence by engaging through restriction or participating in the rapidly developing global ecosystem.
Also Read: Claude Mythos: Imminent Threat or Marketing Hype by Anthropic?
