Discussion Papers

Findings from research and studies conducted at the Financial Research Center (FSA Institute) are organized and published as Discussion Papers to stimulate further discussion and comment. Please send comments to: frtc_comments★fsa.go.jp (replace the star sign with @).
The views expressed in the papers are those of the authors and do not necessarily reflect the views of the Financial Services Agency or the FSA Institute.

Discussion Papers (FY 2025)

No.	Title/Author(s)
DP2025-3 (July 2025)	Study on the Potential for the Utilization of Generative AI in Audit Practices and Challenges Related to Audit Quality NOMA Mikiharu Abstract |  Full text (PDF:1,046KB)(Full text is available only in Japanese)
DP2025-2 (July 2025)	Study on the effective conduct of TLPT by financial institutions in line with global development KITAHARA Yukihiko Abstract |  Full text (PDF:1,336KB)(Full text is available only in Japanese)
DP2025-1 (June 2025)	Study on Insurance Agents Systems in Major Countries: Focusing on Large-Sized Shared Insurance Agents NAKAIDE Satoshi, YAMASHITA Tetsuya, DOKI Takahiro, and JUNG Chanok Abstract | Full text (PDF:1,731KB)(Full text is available only in Japanese)

---

Abstract

DP2025-3
Study on the Potential for the Utilization of Generative AI in Audit Practices and
Challenges Related to Audit Quality 

NOMA Mikiharu, Special Research Fellow, Financial Research Center (FSA Institute)

Artificial intelligence (AI), including generative AI, is rapidly advancing technologically and is affecting multiple facets of our society. The adoption, application, and exploration of generative AI is becoming more widespread in the audit industry. The utilization of generative AI and other technology in auditing is expected to contribute positively to improving audit quality issues and alleviate the shortage of audit resources by enhancing efficiency. On the other hand, it is also assumed that there are issues regarding the use of AI that could negatively impact auditors’ performance and compromise audit quality, such as excessive reliance on AI, AI’s “black box” nature, information leakage, and hallucination.

Numerous discussions are taking place among audit firms, industry associations, audit regulators, and international organizations concerning the potential benefits and challenges associated with the utilization of generative AI in auditing. In academia, prior studies have been conducted on themes such as the potential use of generative AI and possible effects that AI would have on audit firms’ recruitment strategies and practices. For example, Law and Shen (2025) examine whether AI can replace human auditors or complement them, and Fedyk et al. (2022) survey the impact of AI on audit quality and efficiency.

This study examines the potential utilization of generative AI in audit practices and associated challenges in terms of audit quality, based on these prior studies as well as by observing developments in the actual utilization practices of generative AI and discussions among and initiatives taken by audit regulators and international organizations.

Keywords: generative AI, audit, and audit quality.

DP2025-2
Study on the effective conduct of TLPT by financial institutions in line with global development 

KITAHARA Yukihiko, Research Fellow, Financial Research Center (FSA Institute)

The threat of cyber attacks on financial institutions is increasing, and the methods of attack are evolving constantly, becoming more sophisticated and shrewd. In order to respond to such increasingly sophisticated cyber attacks, it is important for financial institutions to identify their cybersecurity risk continuously, and to strengthen their response to cyber attacks by taking measures not merely through system improvement, but from the perspective of strengthening their overall response system, including the corresponding behavior and the development of its response processes against cyber attacks.

In the situation of increasing cybersecurity risk, Threat-led Penetration Testing (TLPT) is considered useful as a tool for assessing how effectively financial institutions’ response systems work against potential cyber attacks, for improving their response systems, and for drawing lessons for further improvement. In Japan, the use of TLPT has been increasing, particularly by major financial institutions. In other countries, initiatives related to TLPT are more advanced, especially in Europe and the United States, and relevant legal systems and frameworks for its implementation have also been developing.

This report provides an overview of developments related to initiatives for TLPT taken by financial institutions in other countries, and frameworks for the implementation of TLPT, including rules and regulations and other legal arrangements. Based on a review of developments in other countries, the report discusses how Japanese financial institutions should conduct TLPT and identifies some of the points that need consideration in conducting TLPT.

Keywords: cyber attacks, cybersecurity, Threat-led Penetration Testing (TLPT).

DP2025-1
Study on Insurance Agents Systems in Major Countries: Focusing on Large-Sized Shared Insurance Agents 

NAKAIDE Satoshi, Special Research Fellow, Financial Research Center (FSA Institute)
YAMASHITA Tetsuya, Special Research Fellow, Financial Research Center (FSA Institute)
DOKI Takahiro, Special Research Fellow, Financial Research Center (FSA Institute)
JUNG Chanok, Special Research Fellow, Financial Research Center (FSA Institute)

In recent years, there has been a series of misconduct in the insurance industry in Japan, such as the fraudulent insurance claim case by large-sized shared non-life insurance agents. The Financial Services Agency of Japan (JFSA) submitted a bill to amend the Insurance Business Act to the Diet in 2025, as part of its response to address such misconduct. To date, Japan’s system for insurance solicitation has progressed based on lessons learnt from experiences in, not only such misconduct, but also past incidents. This kind of improvement in insurance systems is also seen in other countries, and the designs of insurance agent systems of other countries that have been developed based on the circumstances of each country would provide useful knowledge when planning for the further development of Japan’s insurance agent system.

In this study, in order not only to respond to the misconduct but also to contribute to deliberating on the insurance industry from a medium- to long-term perspective, we conducted a survey on the insurance agent systems in the United States, the United Kingdom, Germany, and South Korea. We study each country’s regulation of insurance solicitation by agents, such as licensing and registration systems, regulation of internal management and compliance systems, and regulation of conduct, with a particular focus on large-sized shared insurance agents. There are elements that are common between countries, while there are also differences due to the features of the markets and variations in other economic systems behind the insurance industry. Therefore, we need to deliberate carefully on whether it is suitable to apply foreign systems to Japan directly. However, by studying the situations in the above four countries, we have learnt about the global trends and also recognize clearly the distinct features of Japan’s insurance agent system, and we have obtained implications for studying the ideal insurance agent system in Japan. In the conclusion of the report, we highlight some of the findings from the survey that are of particular interest.

Keywords: insurance, insurance agents, Insurance Business Act, and system for insurance solicitation.