Sam Altman Says AI Development Speed Must Not Outpace Safety and Alignment

Sam Altman, CEO of OpenAI, stated that while regulating the pace of artificial intelligence (AI) development does not mean halting technological progress, the rate of AI capability advancement must not surpass safety, alignment, and monitoring standards. Altman explained that OpenAI has temporarily paused certain advanced reinforcement learning to meet safety benchmarks for upcoming model deployments.
OpenAI confirmed in an official statement that it is temporarily suspending some reinforcement learning training for its latest models ahead of deployment to ensure alignment, security, and monitoring frameworks keep pace with new AI capabilities. The company also noted it is strengthening safety research, alignment studies, and model behavior monitoring.
Korean Source
This article is an English localization of a Korean-language crypto news report. Original headline: 샘 올트먼 “AI 발전 속도, 안전·정렬 앞서선 안 돼”