Japan's Digital Agency Selects 7 Domestic LLMs for Government AI Platform
Japan's Digital Agency has selected 7 domestic AI models, including NTT's tsuzumi 2, for its government AI platform.
The Japan Digital Agency announced on March 6, 2026, that it has selected seven domestic large language models (LLMs) for trial use in the government's generative AI environment, "Gennai." With a declining birthrate and aging population leading to a severe labor shortage, the agency has determined that the use of AI is essential to maintain and improve administrative services. By taking the lead in utilizing AI, the government aims to improve the AI literacy of its employees and foster the domestic AI industry.
The seven selected models are: NTT DATA's "tsuzumi 2," CustomerCloud's "CC Gov-LLM," a joint submission from KDDI and ELYZA called "Llama-3.1-ELYZA-JP-70B," SoftBank's "Sarashina2 mini," NEC's "cotomi v3," Fujitsu's "Takane 32B," and Preferred Networks' "PLaMo 2.0 Prime." These models will be trialed in a large-scale demonstration project involving approximately 180,000 employees from all central government ministries and agencies from May 2026 to March 2027.
The selection was made from 15 applications received during a public call for proposals conducted from December 2025 to January 2026, following a document review and performance evaluation tests. The selection criteria included being developed in Japan, having practical performance for administrative work, and ensuring security. Particular emphasis was placed on Japanese language processing capabilities and compatibility with Japanese culture and values.
This initiative aims not only to introduce AI tools but also to promote a transformation of administrative work processes and organizational culture. The Digital Agency expects each ministry and agency to participate proactively in the demonstration and establish a governance system for AI utilization. In fiscal 2027, the government plans to procure the best-performing models on a paid basis for full-scale implementation, based on the results of the demonstration. This is intended to create stable demand for domestic AI and ensure Japan's autonomy in AI development.
AI Newsletter
Get the latest AI tools and news delivered daily
Related Articles
California Governor Signs Executive Order to Set New Standards for AI Procurement
California Governor Gavin Newsom signed a new executive order on AI procurement for the state government, requiring companies to demonstrate how their systems mitigate risks such as bias, privacy violations, and illegal content, reinforcing the state's leadership in AI regulation.
Anthropic Sues Pentagon Over 'Supply Chain Risk' Designation
Anthropic files federal lawsuit against the Pentagon, challenging its designation as a 'supply chain risk' after refusing AI use in autonomous weapons and domestic surveillance.
Pentagon Designates Anthropic a Supply Chain Risk Amid Dispute Over AI's Military Use
The Pentagon has labeled AI firm Anthropic a "supply chain risk" after it tried to restrict military use of its Claude AI model. The military will phase out Claude over six months, highlighting the conflict between AI ethics and national security.