OpenAI recently confirmed that it will participate in the model evaluation mechanism required by the latest artificial intelligence executive order signed by U.S. President Donald Trump. According to this executive order, AI companies must provide relevant access rights to the federal government before officially releasing new models, enabling the government to assess their capabilities and potential risks. George Osborne, OpenAI’s Global Affairs Head, stated during an interview with media at the SXSW conference in London that the company would join this voluntary framework. He argued that governments of democratic nations should play a significant role in the usage and deployment of artificial intelligence technologies. Osborne noted that as a developer of frontier AI models, OpenAI has always placed great emphasis on safety responsibilities. “We do not wait for government mandates to act; instead, we proactively communicate with regulators and propose ways to track and manage AI safety and risk issues,” he said. This extends beyond the United States to include global cooperation. Under the executive order signed by Trump earlier this week, participating enterprises need to open assessment channels to the government approximately 30 days prior to model release and undergo related benchmark tests. The testing focus will be on evaluating advanced cyberattack capabilities and potential risk levels of AI models, determining whether they should be classified as "frontier models" requiring intensive regulation. Regarding future directions for AI regulation, Osborne, who served as UK Chancellor of the Exchequer from 2010 to 2016, emphasized that governments must maintain sufficient professionalism and foresight when formulating regulatory policies. He suggested establishing technically capable regulatory bodies while allowing flexible adjustment space for future technological developments to avoid overly rigid frameworks hindering innovation. In recent years, with rapid advancements in generative AI capabilities, balancing technological innovation with security assurance has become a core issue shared globally among regulatory agencies and tech corporations. OpenAI's statement indicates leading AI firms are gradually strengthening collaboration with governmental departments regarding model safety assessments and risk governance.