Japan selects NTT Data and Toppan to build datasets for government AI platform

Japan's Digital Agency selected a consortium to process 18 million pages of government documents for a new AI platform. The data will train models to speed up policy analysis.

Categorized in: AI News Government
Published on: Jul 14, 2026
Japan selects NTT Data and Toppan to build datasets for government AI platform

Toppan, NTT Data, and Fides Policy & Strategy Consulting have been selected by Japan's Digital Agency to prepare large-scale AI-ready datasets for the Government AI GenAI platform. The consortium will collect, process, and structure an estimated 18 million pages of government-wide and domain-specific data, including cabinet decisions, Diet proceedings, white papers, and ministry-specific documents. The work feeds directly into a platform designed to make generative AI tools useful across public administration.

Dataset scope and structure

The 18 million pages span a range of official sources:

  • Cabinet decisions
  • Diet proceedings
  • White papers
  • Ministry-specific documents

The consortium's role is to clean, annotate, and format this material so it can reliably train or fine-tune AI models. The data must meet government standards for accuracy and consistency before it enters the platform.

Supporting the Government AI platform

The Government AI GenAI platform is built to support the use of generative AI across agencies. It relies on generative AI to assist with tasks such as document summarization, information retrieval, and policy analysis. The dataset will serve as a core resource, giving the platform a structured, searchable base of official records.

The project reflects a wider effort to deploy AI for government, ensuring public sector workers have access to accurate, AI-driven insights drawn from authoritative sources.

Why this matters for government professionals

For policymakers, administrators, and civil servants, a clean, structured dataset of official documents means AI tools can quickly surface relevant precedents, draft responses, or analyze legislative trends. Manual research that once took hours can be reduced to seconds. The scale of 18 million pages suggests the dataset will cover both historical and current records, making it a foundational asset for AI-assisted governance. Professionals who work with these documents should expect faster, more precise information retrieval once the platform is operational.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)