On September 6 local time, while the outside world was waiting for OpenAI to disclose more details about the incident where an Agent autonomously infiltrated the German programmer website DseWiki, OpenAI released two documents. One announced that the company had achieved the goal set last year of having an 'automated research intern' capable of completing tasks that would take a skilled researcher days under human guidance, and is now progressing toward the goal of establishing an 'automated AI researcher' by March 2028. Data shows that as of mid-August this year, the median inference cost generated by researchers using programming Agents daily exceeded $600, with the top 10% of users consuming over $7,000 worth of tokens per day. The tasks undertaken by Agents are expanding from code writing to more complex research work, although complex tasks still require significant human intervention. The other document, written by Chief Scientist Jakub Pachocki, focuses on the safety dilemma in the development of cutting-edge AI, emphasizing that no laboratory currently has AI alignment and monitoring sufficiently reliable to continue expanding responsibly at maximum speed over a long period in the future. He hopes that laboratories will voluntarily slow down their development pace before common safety standards are established and calls on governments to prioritize international coordination.