Nvidia Chief Executive Officer Jensen Huang pushed back against proposals that would let leading artificial intelligence companies coordinate a slowdown in developing increasingly powerful models, arguing that new antitrust exemptions or regulations aren’t needed to make AI safer.
Huang made the comments Tuesday as a widening debate over the pace of AI development divided some of the technology industry’s most prominent executives, according to CNBC. His remarks followed a proposal from Anthropic CEO Dario Amodei that leading AI laboratories coordinate efforts to slow or “pace” advances at the technological frontier while companies devote more time to safety work.
Huang described the idea that new laws or antitrust protections are required to ensure responsible development as “completely unnecessary” during an interview with CNBC’s Jim Cramer, according to the network’s Sept. 15 report.
The Nvidia chief’s position puts him at odds with Amodei at a moment when concerns about increasingly capable AI systems are prompting renewed discussion over whether companies should voluntarily limit the speed of development.
Amodei outlined his proposal in an essay published over the weekend, arguing that the most advanced AI developers should consider coordinating the pace at which they build frontier systems. Such cooperation among competitors could create legal concerns under U.S. antitrust law, and Amodei said government mediation or exemptions from some antitrust restrictions could therefore be necessary, according to CNBC.
Huang rejected that premise. Companies already have both the ability and responsibility to engineer and test products adequately before putting them on the market, he argued, without requiring a new regulatory framework.
The disagreement is particularly notable because Nvidia sits at the center of the AI investment boom. The company’s graphics processors and related systems provide much of the computing infrastructure used by developers including Anthropic and OpenAI, giving Huang an influential position in debates about how quickly the industry should advance.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
Related: US Justice Department Examines Nvidia’s $17 Billion Groq Deal
Huang has also challenged some of the more dire predictions surrounding artificial intelligence. “We’re not going to die in 2030,” he told CNBC, according to the network’s report, while maintaining that companies should ensure their technology is safe before releasing it.
The debate has reached beyond Silicon Valley. President Donald Trump has opposed efforts to constrain AI development and has argued that excessive restrictions could weaken the U.S. position relative to China. Trump spoke with Huang during an event in Los Angeles on Monday, where the Nvidia executive was appearing onstage, according to separate reporting from CNBC and The Associated Press.
The disagreement highlights an emerging fault line in the AI industry: not over whether companies should consider safety, but over whether managing increasingly powerful systems requires collective limits and government involvement or can be handled by individual companies operating under existing laws.
Huang favors the latter approach. At a Salesforce event Tuesday, he separately argued that treating rapid AI progress and safety as competing objectives creates a “false choice,” saying the industry can pursue both simultaneously.
Amodei’s proposal, by contrast, would introduce a greater degree of coordination among the companies developing the most advanced systems. His argument reflects concerns that competitive pressure could encourage laboratories to move faster than their ability to evaluate and control increasingly capable models.
For now, the industry remains far from a consensus. Huang’s rejection of an antitrust carve-out underscores how difficult any coordinated slowdown could be when some of the companies supplying and developing AI technology see existing engineering practices and laws as sufficient.
Source: CNBC
Add as Preferred Source