2026.07.08
PRESS RELEASE
How Oarai Town Took a Practical Approach to Generative AI Training
On June 22, 2026, Akatsuki AI Technologies conducted a hands-on generative AI training session for staff members of Oarai Town in Ibaraki Prefecture.
As generative AI becomes increasingly widespread, interest in its use is also growing rapidly among local governments. At the same time, many employees still feel that while they understand what generative AI can do, they are unsure how to apply it to their own work, or that although they have tried it once, they were unable to connect it to their day-to-day tasks.
For AI to become embedded in daily operations, simply learning how to use the tools is not enough. What matters is being able to clearly imagine how AI can be applied to one’s own work.
With this in mind, Akatsuki AI Technologies conducted a hands-on generative AI training session for Oarai Town staff. Rather than focusing solely on acquiring knowledge, the training was designed to leave participants thinking, “I want to try this in my work tomorrow.”
Rather than simply explaining how generative AI works and what it can do, the training focused on helping participants see AI as a practical and accessible tool for their everyday work.
To demonstrate this from the outset, the training materials themselves were created using the generative AI tool Claude. The session began by telling participants, “These materials were also created using AI.” The aim was not only to show them the finished materials, but also to give them a tangible sense of how quickly AI can produce work of this level.
We also introduced a basic framework for prompt design to help users communicate instructions effectively to generative AI, organizing prompts around four elements: “Role,” “Context,” “Request,” and “Constraints.”
Using generative AI effectively does not require specialized knowledge. What matters is how users think about “what to ask” and “how to ask it.” By demonstrating this approach in practice, the training was designed to help participants take their first step toward using AI independently.
The training was structured around three steps: “Learn,” “See,” and “Try.”
First, we introduced the characteristics of leading generative AI tools, including ChatGPT and Claude, along with examples of how AI is being used by local governments. This helped participants understand the role generative AI can play in improving administrative services and day-to-day operations.
Next, we demonstrated several use cases based on actual types of municipal work. These included guidance for procedures at bereavement service counters, assistance in searching for subsidy information, analysis related to Japan’s hometown tax donation program (Furusato Nozei), and tourism information. By using themes closely related to real-world operations, we demonstrated how generative AI could support municipal work.
All demonstrations were created using only publicly available information and dummy data. No actual resident information or confidential information was used. This allowed participants to envision practical applications while maintaining appropriate information security considerations.
In addition to showing how generative AI can support tasks such as writing, organizing information, and generating ideas, we also explained fundamental principles for using AI safely in local government. These included “Do not enter personal information,” “AI-generated answers must always be reviewed by a person,” and “Final decisions must be made by staff members.”
One message remained consistent throughout the training: to make effective use of generative AI, it is important to understand what to communicate to AI and how to communicate it.
Generative AI is not万能, but by clearly communicating the purpose, context, and expected outcome, users can increase the likelihood of obtaining outputs that are useful in real-world work.
During the training, we introduced a simple three-step approach: “Open AI,” “Ask,” and “Rephrase.” This was designed to encourage participants to experiment with generative AI more casually as part of their daily work.
Participants also actively exchanged views and asked questions on practical topics directly related to their work, including document creation, image generation, and copyright. The session provided an opportunity to share concrete perspectives on how generative AI can be incorporated into actual operations.
The value of generative AI lies not simply in introducing the technology, but in ensuring that it is continuously used in day-to-day operations.
Akatsuki AI Technologies provides hands-on AI training tailored to the operations and challenges of individual companies and local governments. We also provide end-to-end support, including the development of AI usage guidelines, proof-of-concept (PoC) initiatives, AI system development, and ongoing support following implementation.
Through the development of people who can effectively apply AI in real-world operations, we will continue to support digital transformation and productivity improvements across companies, local governments, and public-sector organizations, while helping establish environments in which AI becomes firmly embedded in everyday work.
