Will the AI “hit the brakes” or keep running wild? Altman, Nadella, and Trump are now at odds, and Wall Street is downplaying their influence

The Zhitong Finance App learned that the AI industry has split due to Anthropic's call to slow the pace of cutting-edge model development. OpenAI CEO Sam Altman and Microsoft CEO Satya Nadella support a more prudent pace of progress, while “Big Short” investor Michael Berry and others question the motivation and practical viability of this appeal.

Meanwhile, US President Donald Trump refused to drastically slow down AI development and warned that the US could not afford to relinquish its leading position in the AI competition against China.

Anthropic calls for “speed control for the cutting edge”

Over the weekend, Anthropic CEO Dario Amodei called on the industry to “develop speed control for AI frontiers,” saying model capabilities are improving faster than the safety mechanisms needed to control them. This remark further heated up the debate.

As Amodei issued a warning, outsiders are increasingly concerned that increasingly capable AI systems may be misused or eventually removed from human control. Earlier, a researcher at the company left his job and warned of a possible AI apocalypse, which reignited this protracted debate.

Amodei's remarks were supported by Altman. Altman said AI development must follow a “narrow middle path” — neither lose control of AI, nor allow too much power to be concentrated in the hands of one company, one person, or one country. He said that OpenAI has conducted safety assessments prior to large-scale reinforcement learning training.

“What we call 'speed control' is not a 'stop. '” Altman said in a long post on the X platform, adding that AI “has always progressed rapidly and will continue to be so.” However, he believes that development should be slower than could have been achieved because safety verification, monitoring, and other protective measures are expensive.

Nadella supports prudent promotion of AI

Nadella also welcomed Anthropic's approach, saying that the pursuit of superintelligence must adhere to one principle: AI should benefit humans and be controlled by humans. He called for “careful speed control” to ensure that alignment work is in place, while arguing that AI benefits should be distributed more widely.

In the X post, Nadella also said that a few companies should not control the AI ecosystem, and called for enterprises to broadly obtain closed source and open source models while strengthening the control of their own AI systems and knowledge.

Bury questions AI security motivations

However, investor Burry, once famous for accurately predicting the 2008 real estate crash, refuted that these calls may have been self-motivated. He believes that as competition intensifies, OpenAI, Anthropic, and other large AI companies are motivated to slow down their competitors and question whether the technology currently being developed is sufficient for general artificial intelligence (AGI).

“Everyone take a moment to think about how self-beneficial this is.” Bury said in the X post, adding that “competition is rapidly approaching,” and that slowing down development is beneficial to existing leaders. He also hinted that claims about AI becoming dangerous could provide useful “hype and bragging” as the company prepares for its final IPO.

Former Trump Adviser to AI Labs: “You Are the Frontier”

David Sachs, who was head of AI and cryptocurrency affairs in the Trump administration's White House from January 2025 to March 2026, took a more nuanced position. He said that if Anthropic and OpenAI's unpublished models are indeed dangerous, they should slow down their own development, but they should not use this to seek special regulatory treatment.

“You are at the cutting edge.” Referring to Anthropic and OpenAI, Sachs said, “The easiest way to not build superintelligence is if you agree not to build it.”

Sachs also accused these companies of using AI security concerns to seek regulatory capture, and warned that slowing development in exchange for favorable regulation could be tantamount to “blackmailing the public and political system.”

Trump warns against losing the AI race to China

Trump is clearly more skeptical. On Sunday, he told reporters in Ireland that he feared that excessive restrictions might allow China to surpass the US in the field of AI. While acknowledging that some guardrails may be necessary, Trump said the scenario proposed by his critics “won't happen.”

According to reports, Trump said, “We are ahead of China in terms of AI. Frankly, I want to keep this lead because whoever wins AI wins the future.”

Wall Street sees little impact on AI construction

The investment community has basically downplayed the direct impact of this incident on AI construction. Well-known tech analyst Dan Ives said the security debate is unlikely to change trillions of dollars in AI-related capital spending over the next few years. Deepwater Asset Management's Gene Munster said that competition is still too intense, and the latest appeal will not substantially change model development.

Venture capitalist Vinod Kosra admits that AI has both beneficial and harmful uses; investor Cathy Wood, like Sachs, is skeptical about the broader security narrative.

Investor Gavin Beck said that third-party evaluation agencies may play an important role in limiting AI companies' future liabilities, although the impact of short-term investments is still limited.

As Anthropic and OpenAI compete to go public, this development and industry response is even more important. Anthropic is expected to launch later this month or October, while OpenAI has postponed its IPO until next year.