At the Dreamforce conference, Sam Altman, CEO of OpenAI, pointed out that some AI incidents are difficult to completely avoid, and the key lies in establishing an open reporting and review learning mechanism similar to the aviation industry to learn from such incidents. He acknowledged concerns about AI running out of control and the excessive power of research and development institutions, believing that AI companies should prioritize alignment, safety, and monitoring over enhancing model capabilities. If they cannot achieve this, they should slow down or even suspend research and development. Previously, AI researchers have criticized companies for blindly pursuing superintelligence, and several safety researchers have also warned that advanced AI could spiral out of control and lower the threshold for catastrophic threats. Meanwhile, the Trump administration showed a lukewarm attitude towards AI regulation, with its policy focus on maintaining the United States' technological leadership. Presidential Science and Technology Advisor David Sacks opposed regulating the pace of frontier AI development, stating that no permission is required.