Tadao Nagasaki, the president of OpenAI Japan, said that artificial intelligence must never be allowed to make final judgments [1].
This caution comes as businesses globally integrate generative AI into core operations. If AI is permitted to make autonomous decisions without human oversight, it could lead to systemic errors or ethical failures in corporate governance.
Nagasaki spoke during a summer seminar organized by the Japan Association of Corporate Executives [1]. The event focused on the future direction of AI and the strategies businesses should adopt to handle its implementation [1].
During the discussion, Nagasaki said that the nature of human labor is evolving. He said that human work will likely shift toward the act of judging. However, he emphasized that while AI can assist in the process, the final decision must remain a human responsibility [1].
Nagasaki said that companies must focus on how to cultivate personnel who are capable of making proper decisions within the organization [1]. This suggests a shift in corporate training, moving away from technical execution, and toward critical thinking and oversight.
Other industry leaders participated in the seminar to discuss the trajectory of the technology. Akio Yamaguchi, the president of IBM Japan, said the current high level of interest in AI and questioned the direction the technology would take moving forward [1].
Similarly, Tadao Kikuchi, chairman of Royal Holdings, said that a new perspective is necessary for the workforce. He said that the goal should be to utilize AI in a way that allows humans to concentrate on creating value [1].
Together, these leaders highlighted a consensus that AI should serve as a tool for efficiency rather than a replacement for human accountability. The shift toward a judgment-based economy requires a workforce that can audit AI outputs rather than simply accepting them [1].
“Ultimately, AI must never be allowed to make judgments.”
The insistence by the head of OpenAI's Japanese division that AI remain excluded from final decision-making signals a strategic boundary between operational efficiency and executive authority. By framing the future of work as a shift toward 'judgment,' the leadership is suggesting that while AI can process data and suggest paths, the legal and ethical liability of a decision must remain with a human. This creates a new imperative for corporate training, where the value of an employee is measured by their ability to override or validate an AI's suggestion.
