Altman issued a public warning, stating that the development of AI has arrived at a perilous juncture. He urged leading laboratories to voluntarily decelerate their research and development (R&D) efforts to guarantee that AI capabilities do not outpace safety alignment and monitoring mechanisms. Altman dissected the two critical threats posed by AI advancement: firstly, there's a risk that humans could lose control and become subordinate to AI; secondly, a high concentration of power might pave the way for a dystopian society. He champions a middle-ground approach, eschewing Silicon Valley's competitive credo of "speed is paramount." Instead, he underscores that deceleration does not equate to stagnation but rather ensures that technological progress is slower than unchecked recklessness. He firmly believes that the cost of implementing safety measures is justified. Currently, OpenAI is spearheading changes by conducting proactive safety evaluations and actively welcoming external oversight. Previously, industry leaders, including the CEO of Anthropic, Musk, and over 1,300 researchers, have advocated for a slowdown in AI development due to safety concerns, such as autonomous escape and the exploitation of zero-day vulnerabilities, which surfaced during mid-2026 frontier model tests. Altman calls for industry self-regulation while also looking ahead to future international coordination and cooperation, stressing that humans must retain control over AI.