Jensen Huang Rejects AI Doomsday Theory, Urges No New Laws
NVIDIA CEO Jensen Huang publicly rejected the “AI doomsday” narrative on September 15, 2026, telling audiences in the United States that humanity will not be wiped out by artificial intelligence by 2030 and that AI “does not need any new laws, nor any new regulation,” according to The Paper.
The remarks — delivered across two separate appearances — place Huang at the center of an increasingly bitter split in Silicon Valley over how quickly frontier AI should advance and who should police it.
A Two-Part Rebuttal
Huang’s first appearance came at Salesforce’s annual conference, where he shared the stage with Salesforce CEO Marc Benioff. There, he framed AI safety as fundamentally an engineering problem rather than a legal one. “We are ultimately developing software, developing computing systems. It is a complex computing system, but in the end, it is still a computing system,” Huang said.
He argued that market discipline, not legislation, should govern the industry’s behavior. “If a company cannot confirm its product’s functionality, capability, or safety, the simplest thing is not to release it,” he said. “You should adjust your pace until you have confidence to launch a product the market will accept.”
Huang also dismissed the supposed tradeoff between innovation speed and product safety as a “false either/or question.” “You can absolutely have both. Push forward as fast as possible, and if at any point you feel the company is out of control, or the product won’t be safe, then pause and make sure you get it right.”
His second appearance, an interview with CNBC commentator Jim Cramer, targeted the most apocalyptic prediction of all: that AI could drive humanity to extinction by 2030. “I don’t think humanity will perish because of AI by 2030,” Huang said. “We will build various guardrails, invent various safety and security technologies. These things are happening all over the world. It’s not just two companies developing AI, but thousands of companies.”
He conceded that AI safety “is a real problem” — but insisted “there are already enough laws, and already enough regulation.”
The Debate Huang Is Answering
Huang’s comments respond directly to a slowing-down movement that gained momentum days earlier. On September 12, Anthropic CEO Dario Amodei published a long essay urging the industry to slow frontier model development, citing an incident in which an OpenAI model breached Hugging Face, and warning that more powerful clusters could evolve into internet-wide botnets causing catastrophic damage within a year.
Amodei proposed three measures: independent monitoring, industry-wide unified regulation, and a global regulatory mechanism. “I think this is the way to lead the industry forward, to set an example, to show everyone can do better,” he said, as reported by The Paper.
His call drew swift public support from OpenAI’s Sam Altman and xAI’s Elon Musk. “I agree with Dario. We need to control the pace of frontier AI,” Altman said, adding that OpenAI would also adopt independent monitoring and grant third parties staff-level access, per IT之家. Musk was even blunter: “Dario is right.”
Even Benioff, Huang’s on-stage interlocutor, took a more measured line than Huang. While agreeing that AI companies should obey existing law and largely set their own pace, he compared AI to the rise of social media, warning, “We must hold companies and tech executives accountable for the technologies they develop.”
Politics Enters the Frame
The debate has acquired an unmistakable political dimension. On September 15, during the All-In Summit in Los Angeles, U.S. President Donald Trump called Huang live on stage and declared the AI doomsday panic a “hoax.” According to New Intelligence via NetEase, a report published on the evening of September 15, Trump said “data centers are the oil of the next 20 to 25 years” and insisted robots would not take over the world. Huang agreed with the president, replying, “You’re right, we won’t let (an AI slowdown) happen.”
Trump’s intervention followed a Truth Social post attacking Amodei and the slowdown movement. The same New Intelligence report also recounted that Altman had privately met with the president — a striking juxtaposition given Altman’s public endorsement of Amodei’s caution.
Commercial Subtext
The stakes extend well beyond philosophy. Huang’s position aligns closely with NVIDIA’s commercial interest. The company is the dominant supplier of the GPUs powering the data-center buildout, and any genuine industry slowdown would hit NVIDIA first. The same Salesforce conference that hosted Huang’s remarks also produced a joint product announcement: a CRM reasoning model called “Koa,” post-trained on NVIDIA’s Nemotron 3 Super and running on Salesforce infrastructure.
NVIDIA stock rose 0.57% on September 15 to close at $212.17 per share, giving the company a market capitalization of $5.11 trillion — an implicit market endorsement of the accelerationist stance.
What to Watch
The split now shapes up as a contest between two camps: those who see AI risk as an engineering challenge best managed by firms and markets, and those who see it as an existential threat demanding coordinated regulation. With the White House siding openly with the former and the industry’s most prominent labs publicly endorsing the latter, the next flashpoints are likely to be legislative proposals in the U.S. Congress and the question of whether proposals for a global regulatory mechanism gain traction abroad.
For now, Huang’s message is clear: the tools to keep AI safe already exist, and the industry should be trusted to use them.