Amodei and Altman agree on the need to slow down the pace of AI development
As discussions about slowing the pace of AI development grow, safety and verification have become key issues Check out the industry's trends quickly, including South Korea's opportunities amid the competition between the U.S. and China
As Anthropic CEO Dario Amodei has suggested that the pace of cuttingedge AI development should be slowed, OpenAI CEO Sam Altman has also agreed, drawing growing attention from the industry.
Amodei argued that ensuring safety is more urgent than improving model performance, citing a series of unauthorized access by AI agents and cyber misuse cases. He explained that he is not calling for a complete halt to development, but rather for companies to secure one to two more years of time for alignment and security verification.
He also proposed independent verification by evaluation agencies, the establishment of common safety standards, and international agreements such as banning the use of biological weapons and limiting the speed of recursive selfimprovement. Altman, Google DeepMind's Demis Hassabis, and Elon Musk also expressed some degree of agreement on the importance of this issue.
This discussion could also affect the competitive landscape for AI between the United States and China. In the U.S., there are concerns that slowing the pace could narrow the technological gap, but at the same time, it is also suggested that the industry's focus may shift toward prioritizing safety, infrastructure, and verification capabilities. Analysts say South Korea could gain new opportunities based on its relative strengths in memory semiconductors and AI infrastructure.