The Tokyo Metropolitan Government will begin using the artificial intelligence chatbot ChatGPT for writing texts and carrying out other clerical work in all of its offices from August, Gov. Yuriko Koike said Tuesday.

ChatGPT "has the potential to greatly transform the way public administration is conducted," Koike said during a metropolitan assembly session. She added that "better city governance" can be achieved by assessing the positive and negative aspects of the AI service.

Tokyo Gov. Yuriko Koike | KYODO
Tokyo Gov. Yuriko Koike | KYODO

Koike also said the metropolitan government will use ChatGPT for tasks including preparing documents in question-and-answer format, and seek input from its employees about other practical uses for the generative AI tool.

Chatbots are software applications trained using massive amounts of data from the internet, enabling them to process and simulate human-like conversations with users. It gives responses based on users' instructions and questions.

To mitigate concerns, such as preventing leaks of confidential information, the metropolitan government has set up a project team to test the efficiency of ChatGPT and create guidelines for its use, according to its officials.

Last week, Yokosuka, Kanagawa Prefecture, began using the chatbot. A one-month trial from April, the first by a local government in Japan, showed it improved work efficiency for tasks such as composing texts.

If the use of ChatGPT continues, working hours can be reduced by at least about 10 minutes a day, the Yokosuka government estimated.