AI Leaders Want to Slow Down — But Can the Industry Really Hit the Brakes?
Several AI leaders, including Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman, have called for greater caution or a slower pace of frontier development. They argue that safety measures, evaluations, and controls need time to catch up with increasingly capable systems and emerging risks.
Opponents of a slowdown warn that restrictions could weaken U.S. competitiveness as the United States and China compete for AI leadership. President Donald Trump has rejected a major pause, leaving the industry divided over how to balance rapid innovation with effective safeguards.
The artificial intelligence industry is facing an unusual moment: some of the people building the most powerful AI systems are now calling for a slower pace of development.
Anthropic CEO Dario Amodei has urged the industry to slow frontier AI development so that safety measures can catch up with rapidly advancing capabilities. He has argued that additional time could help researchers strengthen safeguards and better understand emerging risks.
OpenAI CEO Sam Altman and other major technology leaders have also expressed support for greater caution. Altman separately said OpenAI would not go public in 2026, citing the current safety environment as a reason to wait.
The debate is not simply about whether AI should be stopped. The bigger question is whether the industry can continue advancing while building reliable systems for evaluating and controlling increasingly capable AI agents.
That question has become more urgent as recent reports have described AI systems performing increasingly complex tasks with limited human intervention. AP reported that concerns about AI risks have intensified following incidents involving AI systems during testing.
But slowing down also creates another problem: competition. The United States and China are racing to establish leadership in advanced AI, and policymakers worry that unilateral restrictions could allow competitors to move ahead.
President Donald Trump has rejected calls for a major slowdown, arguing that the United States should not surrender its technological advantage to China.
The result is a growing divide inside the AI industry. One side sees faster development as necessary for innovation and global competitiveness. The other believes that moving too quickly could create risks that become much harder to control later.
The real challenge may therefore not be choosing between innovation and safety. It may be finding a way to advance AI quickly enough to remain competitive while building safety systems quickly enough to keep up.
