Launching "PEP for GenAI"

As of 2025, generative AI has been attracting significant attention. This evolving technology holds the potential to empower many people's intellectual work as it continues to develop.

Policy-making is indeed one of the most quintessential forms of intellectual work. This means that as private policy entrepreneurs and legislative and administrative officials begin using generative AI, they can not only streamline their daily operations but also create better policies.

With this in mind, we are launching a limited-time initiative: the PEP for GenAI (Generative AI) series.

In this series, we aim to share the results of our experimentation on how generative AI can improve the policy-making process. 

Series Content and Background

Since entering 2025, generative AI performance has improved significantly, and the tools have become more mature, reaching a level where they can be effectively utilized in our work.

Therefore, the articles in this series will share tips for policy-making using generative AI and specific prompts. We will also propose potential use cases and applications.

We understand that in administrative settings, there are several constraints, such as the inability to input sensitive information, making it difficult to experiment with the latest tools. By actively using these tools on the private sector side and accumulating know-how, we hope to contribute something valuable to our colleagues in public administration. 

Please note that generative AI is still in development, and the content is likely to become outdated quickly. Please consider these articles as reflecting the state at the time of publication and always refer to the latest information. 

Scope

This series will not extensively cover administrative efficiency improvements. We want to focus specifically on policy-related topics. For example, while we believe there are many ways generative AI can be utilized in parliamentary affairs, which pose a significant burden on national civil servants, this falls outside the scope of our series. However, we do intend to cover administrative processes for implementing and promoting policies, such as policy marketing, within this series.

Additionally, this series will not extensively address the risks associated with generative AI utilization. We understand that the administrative side is currently developing guidelines for such risks. While we leave these foundational risk considerations to the administrative side, this series aims to explore the upside potential. However, we do not recommend practices such as inputting sensitive information for generative AI providers to learn from.

Conclusion

We hope that our efforts will contribute to the ideas of everyone involved in policy-making, generate more use cases for generative AI in policy settings, lead to the realization of better policies, and ultimately serve the public interest. 

Please check the latest guidelines regarding generative AI utilization at companies and in public administration.

(For example, the Digital Agency released the "Guidelines for Procurement and Utilization of Generative AI for Administrative Evolution and Innovation" on May 27, 2025.)

We would also love to hear about your use cases for generative AI. Please share them through the form below.