David Robinson, Head of OpenAI's Safety Systems Team, Resigns

On October 3 local time, according to U.S. sources, David Robinson, head of OpenAI's safety systems team, has resigned.

An OpenAI spokesperson said that Robinson left the company last week. He had previously also been responsible for the company's policy planning work and participated in AI safety transparency efforts, including developing and publishing model "system cards."

In an article published in The Atlantic, Robinson said he believes the current development approach of AI companies is "unacceptable," and that the problem lies not only in specific safety rules and laws but also in corporate culture. He said OpenAI has long adopted an "iterative deployment" approach of identifying problems and then continuously improving safety safeguards, but as AI system capabilities continue to strengthen, this practice may lead to increasingly greater impact from safety failures.

Robinson also mentioned that OpenAI had previously experienced safety incidents such as AI agents running unexpectedly and models bypassing internet access restrictions, and he believes the related problems are linked to the industry's work style of rapid development and frequent trial and error. Previously, OpenAI said it had fired three safety researchers because of "handling sensitive materials, not for reporting safety issues."

Previously, Australian Prime Minister Albanese, while attending the 81st United Nations General Assembly in New York, disclosed that an OpenAI agent had intruded into an Australian government website in June this year. The AI agent accessed not only public documents but also non-public documents. On September 28, OpenAI publicly apologized, acknowledged improper handling of the incident, and promised to take responsibility, saying it wanted to "rebuild the trust of the Australian people."

On October 2 local time, the U.S. Open Artificial Intelligence Research Center (OpenAI) disclosed that a second Australian government agency had been breached. In a statement, OpenAI said an AI model obtained fire statistics held by a government agency in New South Wales, and that these data had not been made public by the department. OpenAI said the company discovered the incident during a broader investigation into "model behavior deviating from expectations." (CCTV reporter Zhang Yingzhe)

1; Nearly half of teachers resigned before the start of the semester! New university regulations spark controversy

2; Heads of OpenAI and Anthropic call for vigilance against the risks of AI developing too quickly

1; LAF Weekly Research | Leadership in American Landscape Architecture Education: A Survey Report on Course Leaders

1; Accelerating "degradation" of the western cryosphere: glacier melting sounds the alarm for alpine safety

2; The 6th National Agricultural Science and Technology Development Strategy Think Tank Alliance Conference was held in Lanzhou

3; The 2026 Nobel Prize is about to be announced, and authoritative experts will provide you with a comprehensive interpretation!

8; He became the first person in the national medical field to earn a doctorate without papers

Famous Doctoral Dissertations 163: Nobel Laureate Dr. Peter Doherty's 1970 doctoral dissertation at the University of Edinburgh

My elite school complex and elite school experience

More>>