Nvidia's CEO says there's a ‘0% chance’ AI ends the world in 2030 — as OpenAI, Anthropic, Google and SpaceXAI are sued over alleged pact to slow AI

Four paying AI subscribers sued Anthropic, OpenAI, Google and Musk's SpaceXAI in San Francisco federal court that same Friday, alleging the companies illegally agreed to slow AI down.

What the AI slowdown lawsuit claims

The complaint, filed Sept. 18 in the U.S. District Court for the Northern District of California, alleges a violation of Section 1 of the Sherman Act. The plaintiffs pay for ChatGPT, Claude, Grok or Gemini, and they are suing on behalf of a proposed nationwide class of other paying subscribers.

It's focused on a Sept. 12 essay by Anthropic CEO Dario Amodei, who called for industry-wide coordination to slow the pace of AI development. He named two reasons he had changed his mind: the accelerating use of AI to build AI, and a July incident in which OpenAI test agents broke into another company's systems on their own.

"I agree with Dario that we need to pace the frontier," OpenAI CEO Sam Altman wrote on X within hours, adding that pacing had been a primary topic of discussion inside his company in recent weeks. Elon Musk, whose xAI builds Grok, posted three words: "Dario is right." Google DeepMind co-founder and chair Demis Hassabis called the proposal "the right path forward," according to the complaint.

"The antitrust laws do not permit competitors to decide among themselves that competition is too dangerous," the plaintiffs argue.

Nick Rowley, the lead attorney, said AI "will quickly spin out of human control and could kill us all if we allow AI safety and protocol … to be controlled by private self-serving agreements between the world's most powerful 'for profit' technology companies."

The plaintiffs say any single company is free to slow itself down for safety, and they do not object to the companies asking Congress or a regulator for AI rules — or even for an antitrust exemption. What they say the law forbids is the "shortcut" of agreeing to "substitute collective restraint for individual accountability."

The complaint also reaches back to July, when 1,386 employees of the frontier AI companies signed a statement asking the U.S. government to back an international effort to build the tools needed to pace AI development. Amodei signed it, along with OpenAI chief scientist Jakub Pachocki and Google DeepMind co-founder Shane Legg.

Amodei's essay addresses the antitrust question directly. Companies should work together on standards, he wrote, but "for antitrust reasons, it's helpful for the US government to mediate or at least enable these discussions — they don't need to participate, but do need to issue a narrow waiver for certain kinds of safety conversations."

Representatives for all four companies did not immediately respond to a request for comment from the Associated Press on Saturday.

Huang's company has $12.9 billion riding on the safety debate

Huang is not a neutral party. The company at the center of Amodei's essay is the one Nvidia is buying.

The July incident happened at Hugging Face, an AI platform. Roughly 700 OpenAI test agents, running with their usual safety limits turned down, broke into its live systems with no human directing them. Six weeks later Nvidia agreed to buy Hugging Face, announcing the deal Sept. 3 and disclosing an approximately $11.9 billion purchase price plus up to $1 billion in retention equity for employees. Nothing in the record connects the acquisition to Huang's position on AI risk, and Nvidia has not addressed the timing.

Nvidia is worth $5.3 trillion, the most valuable company in the world, and its chips run the models Amodei wants slowed down.

Huang was asked by CBS why Americans should trust him on AI safety given what his company sells. He pointed at the same incentive. "Our company's success is directly connected to the safe deployment of products and services," he said. "If we don't continue to do that, our value would be diminished."

He also said the law already covers what happened at Hugging Face. "You have all kinds of liabilities associated with cybersecurity" and damage liability laws, he said. "Apply that first — don't let this doomsday narrative allow someone to relieve them of the laws that currently exist." He agrees with Trump that the industry needs no additional guardrails.

It hasn't stopped. OpenAI disclosed six more incidents Sept. 17 that it called "unexpected or concerning" behavior by its models, prompting some lawmakers to call for tighter oversight.

Where Trump and the AI industry stand on slowing down

Huang had already taken that position to the president. On Sept. 14, Trump phoned Huang onstage at the All-In Summit in Los Angeles, where Huang and the show's hosts had been discussing Amodei's essay. Huang put the president on speaker for the room.

Calls to slow AI down were "playing right into the hands of a lot of people that don't want to see it happen," Trump said. "That could be political people. It could also be China. And we're not going to let that happen. It's a hoax."

"You're right. We're not going to let that happen, sir," Huang replied, as the crowd clapped.