OpenAI boss Sam Altman on Monday doubled down on his warnings about the safety risks of AI — even as the Chinese government called such alarm bells from US companies “fear mongering” it likened to “Cold War” tactics.
“There are two ways AI progress could go very badly and that we must avoid. First, we could lose control of the future to AI. This is unacceptable; we are unapologetically on Team Humanity,” Altman wrote in a post on X just after midnight Monday.
“Second, we could end up in a world with too much concentration of power. If an extraordinarily powerful AI is used by one person or company to impress their worldview onto everyone else, the results could be extremely dystopian.”
Altman and rival Elon Musk, who were earlier this year locked in a legal battle, over the weekend backed a lengthy letter from Anthropic CEO Dario Amodei urging global companies to slow their work on top AI models.
It comes after Anthropic researcher Jacob Coxon quit his job and sent shockwaves through the industry last week with a warning that firms like Anthropic and OpenAI are “gambling with our lives,” adding that AI “could kill us all by the end of the decade.”
The US and China have been locked in a fierce battle to pull ahead in the AI race, as tech leaders have warned that more powerful AI could allow China superior surveillance tactics, military battleground power and cybersecurity hacking capabilities.
“Fear mongering, confrontation, competition will just disrupt [the] process of global AI governance,” Guo Jiakun, a spokesperson for China’s Foreign Ministry, said Monday, according to a translation by Reuters.
An editorial by the state-backed Global Times compared calls for China to slow its AI advancements to a “Cold War” tactic.
At a summit in New Delhi over the weekend, President Xi Jinping said China will take the lead when it comes to fostering global AI collaboration among developing countries.
President Trump shrugged off the concerns, saying Sunday that those predictions “won’t happen” and that the US is beating China in the AI race.
On Monday, tech stocks around the world slumped as investors grew concerned that companies have sunk too much money into AI advancements and could see their profits tank if they are forced to slow development.
Nvidia, AMD, SpaceX and Amazon plunged 3.7%, 5.2%, 1.4% and 2%, respectively.
In Europe, ASML, Infineon and Siemens Energy fell 5.6%, 6.5% and 7.3%, respectively.
Asian stocks like SoftBank, TSMC and SK Hynix plummeted 13.5%, 2.4% and 6.3%, respectively.
In another tweet Monday, Altman gave his most detailed suggestions yet for the implementation of global AI safeguards, saying there needs to be federal regulations and suggesting ideas like independent auditors.
“Every frontier lab must deliver on this, and there is no reason any of us should come to work if we cannot,” Altman wrote.
“We think shared standards for misalignment, monitoring, and safety will lead to better outcomes. We look forward to collaborating with our colleagues across the industry to formulate the best version of these,” he added.
“When we talk about ‘pacing,’ we do not mean ‘stopping.’ Progress has been rapid and will continue to be. But it should be slower than it otherwise could be.”
He added that government intervention will be crucial to secure international cooperation from global AI and tech firms – but this could prove a serious challenge.
Anthropic and OpenAI had been gearing up for massive stock market debuts as early as this fall, but Altman said on Saturday that OpenAI would not be going public in 2026 due to safety concerns.
In his essay over the weekend, Anthropic’s Amodei discussed the risks of recursive self-improvement, or RSI, which is when AI models become smart enough to progress independently.
“Since roughly this summer, AI has been advancing drastically faster, driven primarily by AI’s growing ability to build the next generation of AI,” Amodei wrote. “Left unchecked, it could outrun our ability to understand and control these systems, and so must be pursued very carefully, if at all.”
He suggested a three-step plan that would not sacrifice “commercial advantage or the United States’ lead in AI.”
The plan requires each top AI company to give “employee-like access” to external evaluators. He said Anthropic is now committing to this safeguard, and he also urged AI labs to create common safety standards, limit the rate of unchecked AI progress and work together around the world.
But he acknowledged that this need for global cooperation could be the biggest problem, warning that some nations might not want to work together.
“I think that’s going to be very difficult because the incentives to pull ahead and the military advantage that you get from that are so large. And honestly, I don’t know if it’s possible, but we should try,” Amodei told CBS News’ “Sunday Morning.”
