Japan’s central government is set to deploy a GenAI platform across all ministries by the end of March, marking a significant step in efforts to reduce the heavy workloads faced by senior bureaucrats.

The AI system, known as Gennai, was developed by the country’s Digital Agency and will be distributed to government workers nationwide. 

Initially adopted on a trial basis by select bodies, including the Digital Agency, the use of generative AI will now be expanded to all ministries and agencies. Officials are also exploring how the technology could be extended to local governments.

One of the primary motivations behind the rollout is Japan’s chronic overtime problem among central government staff. According to the National Personnel Authority, central government bureaucrats average 376 hours of overtime annually — more than twice the 181-hour average recorded among other public-sector employees. A major contributor is the time-intensive process of preparing draft responses to questions raised in the Diet.

Traditionally, officials prepare multiple versions of draft answers for each parliamentary question, anticipating various follow-up scenarios. This preparation often stretches late into the night. Gennai is expected to significantly cut the time required by generating draft responses and suggesting answers to potential follow-up questions, which staff will then review and refine.

Under the new system, ministries will also be able to develop specialized AI agents trained on Diet minutes, lawmakers’ written inquiries, and historical responses from government officials. Because Gennai was developed internally by the state, it is authorized to access certain highly confidential government data — an important distinction from commercial AI tools.

Beyond legislative support, Japan plans to apply AI to administrative processes such as licensing and permitting. A pilot project is already underway to use AI in reviewing applications from companies seeking certification for workplaces that support women and parents of young children.