Tokyo, March 6 (Jiji Press)–Japan’s Digital Agency said Friday it will start in May a large-scale test of its generative artificial intelligence platform for administrative tasks, involving some 180,000 staff workers at all government agencies and ministries. Through the test, the agency aims to ascertain the effects of generative AI utilization and sort out issues related to its use for work style and task process reforms, in the run-up to the full-scale implementation of the system in fiscal 2027. The “Gennai” generative AI platform, developed and operated by the agency, will be used for the test. Specifically, the agency will use it for administrative tasks and to develop an AI app for supporting the preparation of statements and remarks for parliamentary meetings. The agency will also explore the possibility of introducing an AI agent, which operates in coordination with humans. Also, the agency selected domestically developed seven large language models for use in the test, aiming to support LLM development by Japanese companies and research institutions. After assessing the LLMs, the agency will consider purchasing one with superior performance in fiscal 2027. END [Copyright The Jiji Press, Ltd.]
Japan Digital Agency to Start Generative AI Test in May