US AI Leaders Call for Development Slowdown Amid China Race

Leading AI executives advocate for restraint

Senior executives at America's most prominent AI companies are publicly advocating for a deliberate slowdown in frontier artificial intelligence development—a position that would have seemed implausible just months ago. Dario Amodei of Anthropic, Sam Altman of OpenAI, and Elon Musk of SpaceX AI have each recently called for "pacing" advanced AI work, citing concerns that capabilities are outstripping safety measures.

The timing is significant. These statements emerged in the weeks leading up to a summit between US President Donald Trump and Chinese President Xi Jinping scheduled for Thursday in Washington. Over the weekend, US Treasury Secretary Scott Bessent and Chinese Vice-Premier He Lifeng met in New York City specifically to discuss AI safety issues, according to AI Watch.

Why it matters

This represents a fundamental tension in AI governance: US companies face pressure to maintain technological leadership over China while simultaneously managing existential risks from the technology itself. The call for pacing suggests safety concerns may be reaching a threshold where even competitive dynamics cannot justify unchecked acceleration. Whether voluntary restraint can work when geopolitical rivals continue development remains an open question with profound implications for global AI policy.

The competitive paradox

The executives' proposals highlight a core dilemma in AI development. While these leaders express genuine concern about advancing capabilities beyond existing safeguards, their companies remain locked in intense competition—both with each other domestically and with Chinese AI efforts internationally. This creates a prisoner's dilemma where unilateral slowdowns could cede strategic advantage.

The backdrop of US-China relations adds complexity. The two nations previously agreed to work toward "constructive strategic stability," but translating that principle into concrete AI governance remains challenging. The recent meeting between Bessent and He Lifeng signals that both governments recognize AI safety as requiring bilateral coordination, even amid broader technological competition.

From acceleration to caution

The shift in rhetoric from AI leaders marks a notable evolution in industry thinking. After years of emphasizing rapid capability improvements and racing to deploy increasingly powerful models, the conversation has turned toward risk management and deliberate development timelines. Whether this represents a genuine strategic pivot or rhetorical positioning ahead of potential regulation remains to be seen.

The concept of "pacing" itself lacks precise definition—it's unclear whether executives envision formal development moratoria, voluntary industry standards, or government-imposed speed limits on certain research directions. These details will prove critical as policymakers consider how to balance innovation incentives against safety imperatives.

These details were first reported by AI Watch as part of a series examining US-China AI dynamics.