Microsoft CEO Satya Nadella Backs Artificial Intelligence Slowdown as Beijing Rejects Calls to Curb Chinese AI

Nadella said the pursuit of increasingly capable artificial intelligence systems should remain focused on human control while ensuring that the benefits of the technology are broadly accessible. His comments came shortly after Anthropic CEO Dario Amodei called for the development of frontier artificial intelligence to be paced more carefully and urged greater international coordination.

At the same time, Beijing has rejected Amodei’s call for continued restrictions on China’s access to advanced AI chips and semiconductor equipment. Chinese officials argued that confrontation and competition between major powers could undermine efforts to establish global AI governance.

The contrasting positions highlight a growing divide over how governments and technology companies should balance AI safety against the race for technological leadership.

Satya Nadella calls for a measured approach to AI development

In a post on X, Nadella said any pursuit of superintelligence should be based on the principle that artificial intelligence must remain beneficial to humanity and under human control.

“Any pursuit of superintelligence has to be grounded in the core principle that AI remains helpful and controllable,” Nadella wrote.

Nadella calls for human-controlled AI that benefits society while ensuring its advantages are broadly accessible through both closed and open-source models. Source: Satya Nadella via X

He also argued that slowing specific aspects of artificial intelligence development should not come at the expense of wider access to the technology. Nadella called for both closed- and open-source AI models to remain part of the broader ecosystem, saying their benefits should reach countries, communities, and companies across the world.

Nadella’s position is therefore more nuanced than a blanket call to halt AI development. He emphasized deliberate pacing and stronger safeguards while simultaneously advocating broader adoption and competition across the artificial intelligence stack.

The Microsoft CEO also stressed that companies should maintain control over their own knowledge and AI systems rather than becoming overly dependent on a single model provider. He said businesses should have the ability to build continuous learning systems around their own proprietary knowledge and models.

“We welcome the research, focus, and deliberate pacing needed to get alignment right,” Nadella said, while also pointing to mechanisms such as embedded evaluators as potential ways to make AI safety commitments more measurable.

The comments come amid a broader industry discussion triggered by Amodei’s September essay, We Must Pace the Frontier. The Anthropic CEO argued that AI capabilities are advancing rapidly enough to require additional time for safety and alignment work. He proposed independent evaluators, coordination among companies in democratic countries and eventually international cooperation.

Amodei specifically pointed to the possibility of recursive self-improvement, in which AI systems increasingly contribute to the development of more capable systems. He argued that this could accelerate progress beyond the ability of researchers and regulators to adequately evaluate the resulting risks.

His proposal did not call for an immediate halt to artificial intelligence research. Instead, Amodei described “pacing” as a way to ensure that safety measures and independent evaluations keep pace with rising capabilities.

Beijing hits back at CEO’s call to curb Chinese AI

Beijing has taken a different position on the question of slowing AI development, particularly where the issue intersects with China’s technological competition with the United States.

China’s Foreign Ministry responded to Amodei’s argument that Washington should maintain restrictions on advanced AI chips and semiconductor manufacturing equipment reaching China. Amodei said preserving a U.S. lead was important because he viewed a Chinese lead in artificial intelligence as a serious national security concern.

Guo Jiakun called for open and inclusive global AI cooperation, warning that confrontation and threat narratives could undermine efforts to benefit humanity. Source: Global Times via X

Foreign Ministry spokesperson Guo Jiakun urged countries to pursue broader international cooperation instead. He warned that “fearmongering, confrontation, and vicious competition” would disrupt global artificial intelligence governance and serve no one’s interests.

The response reflects the broader geopolitical tension surrounding AI chips, model development, and access to computing infrastructure. Washington has used export controls on advanced semiconductors and related technology as part of its effort to limit China’s access to the most powerful AI computing systems.

Amodei has argued that such restrictions are compatible with a global artificial intelligence safety strategy. In his proposal, he said democratic countries should retain enough of a technological lead to create room for safer development while also seeking limited forms of cooperation with China.

He acknowledged that global AI agreements would be difficult because neither side could easily verify whether the other was complying. His proposal therefore ranged from narrow agreements on particularly dangerous applications to more ambitious arrangements that could eventually place limits on the pace of artificial intelligence advancement.

China, meanwhile, has continued to promote international artificial intelligence cooperation while presenting itself as a participant in the development of open-source AI. Chinese companies have produced models that compete with U.S. systems on several benchmarks and can often be offered at lower costs, adding another dimension to the competition between the two countries.

The artificial intelligence slowdown debate exposes the US-China divide

The disagreement comes as AI development becomes increasingly tied to national security and economic policy.

U.S. President Donald Trump has rejected calls for Washington to slow the artificial intelligence race, arguing that maintaining America’s lead over China is strategically important. Trump said, “whoever wins AI, wins,” underscoring the administration’s focus on technological competition.

Global AI stocks fell Monday after Anthropic CEO Dario Amodei called for slower AI capability gains, with Sam Altman and Elon Musk backing a more measured pace. Source: CNBC via X

That position contrasts with the more cautious approach now being discussed by several AI industry leaders. Amodei has called for stronger safety mechanisms, while OpenAI CEO Sam Altman and other technology figures have expressed support for greater attention to the pace of frontier artificial intelligence development.

Nadella’s intervention adds an important middle position to the debate. His comments support deliberate pacing and human oversight without advocating a broad suspension of AI innovation. He also emphasized competition, accessibility, and the ability of businesses to retain control over their own AI systems.

The issue is likely to remain prominent as Washington and Beijing prepare for further discussions on artificial intelligence governance. Trump and Chinese President Xi Jinping are expected to discuss artificial intelligence governance during a planned September 24 meeting, placing the technology at the center of broader U.S.-China relations.

For the AI industry, the debate is increasingly moving beyond whether development should be fast or slow. The more immediate question is how governments and companies can maintain technological progress while building systems capable of evaluating safety risks, protecting human control, and limiting the consequences of increasingly autonomous artificial intelligence.

Nadella’s comments suggest that the emerging industry consensus may not be a simple choice between unrestricted acceleration and an outright AI pause. Instead, the debate is increasingly focused on whether frontier AI can advance at a pace that allows safety research, independent evaluation, and international governance to keep up.