KAWASAKI, Japan, Feb 3, 2026 - (JCN Newswire) - Fujitsu today announced that it has collaborated with a central government agency in Japan to conduct a demonstration experiment utilizing its Takane [1] large language model (LLM) in public comment operations. The experiment, conducted in 2025, successfully automated advanced tasks such as classifying opinions by support/opposition and summarizing them, demonstrating improved operational efficiency and quality. The agency's officials confirmed the effectiveness of the system. Based on the results of this demonstration experiment, Fujitsu is developing a generative AI service that can be applied comprehensively to policy formulation and legislation, aiming for provision by fiscal year 2026.
In addition to the above, Fujitsu is also planning for the construction of AI workflows that systematically integrate appropriate AI models and tools throughout the entire legislative drafting process, which involves extensive research into related laws and coordination with various stakeholders, and the development of AI agents that autonomously support complex research and coordination tasks arising within these processes. Fujitsu will support government initiatives for AI infrastructure development and generative AI utilization, such as Government AI, and aims to provide a generative AI service by fiscal year 2026 that can be comprehensively applied to policy formulation and legislation processes requiring the robust Japanese language capabilities of Takane.
Results of the Proof-of-Concept (PoC)
In conventional Japanese governmental public comment operations, officials read and classify submitted opinions, analyze trends, draft responses to each opinion, and then consider government policy. This process can take over a month until the results are announced.
In this demonstration experiment, Fujitsu verified the efficiency improvements achieved by utilizing Takane with past public comment data. By applying Takane to approximately 120,000 characters of actual public comment data received by the central government agency that conducted this experiment, tasks such as classifying opinions by support/opposition and summarizing them, which were previously performed manually, were automated and completed in about 10 minutes. This demonstrated the potential for officials to focus on reviewing the output. Furthermore, in checking the consistency between draft laws, which are the subject of public comments, and individual opinions, Takane was able to correctly identify the relevant clauses in the draft law for over 80% of the opinions when both the draft law and opinions were input. This confirmed the potential for significant labor savings compared to manual linking.
These results suggest that by reducing the time administrative officials spend organizing opinions, they can allocate more time to more critical decision-making tasks, such as deliberating the content of opinions and reflecting them in policy.
Utilizing digital technologies, including generative AI, to enhance public comment operations, is also considered an important initiative from the perspective of EBPM (Evidence-Based Policy Making).
Figure: Overview of the PoC
Background
Public comment in rulemaking by national and local governments is a crucial mechanism to ensure that citizens affected by these rules have the opportunity to express their opinions directly. However, for topics of high public interest, administrative officials can receive many thousands of submissions, leading to unmanageable review workloads. This could lead to public opinion not being appropriately reflected in policy and an erosion of public trust.
The central government agency that conducted this demonstration experiment places importance on utilizing digital technology to reduce the burden on officials and improve the quality of public opinion reflection. In collaboration with Fujitsu, the agency undertook this demonstration experiment to apply generative AI to public comment aggregation and analysis, a task highlighted by the Digital Agency in its verification of generative AI utilization.
[1] Takane:
A large language model co-developed by Fujitsu and Cohere Inc.
About Fujitsu
Fujitsu’s purpose is to make the world more sustainable by building trust in society through innovation. As the digital transformation partner of choice for customers around the globe, our 113,000 employees work to resolve some of the greatest challenges facing humanity. Our range of services and solutions draw on five key technologies: AI, Computing, Networks, Data & Security, and Converging Technologies, which we bring together to deliver sustainability transformation. Fujitsu Limited (TSE:6702) reported consolidated revenues of 3.6 trillion yen (US$23 billion) for the fiscal year ended March 31, 2025 and remains the top digital services company in Japan by market share. Find out more: global.fujitsu
Press Contacts
Fujitsu Limited
Public and Investor Relations Division
Inquiries
Copyright 2026 JCN Newswire. All rights reserved. www.jcnnewswire.com