Mitsubishi Research Institute Begins Practical Provision of "AI Scoring Model" to Mebuki Financial Group
Mitsubishi Research Institute (MRI) has started providing its "AI Scoring Model" for housing loans to Mebuki Financial Group. This initiative aims to improve the accuracy of credit risk prediction and support the DX and efficiency of screening operations. Mebuki Financial Group is the first to adopt this model.
📋 Article Processing Timeline
- 📰 Published: April 1, 2026 at 23:00
- 🔍 Collected: April 1, 2026 at 16:47
- 🤖 AI Analyzed: April 21, 2026 at 23:12 (486h 24m after Collected)
Mitsubishi Research Institute, Inc. (Representative Director and President: Kenji Yabuta, hereinafter MRI) began providing its "AI Scoring Model" for practical use to Mebuki Financial Group (Representative Director and President: Tetsuya Akino, hereinafter Mebuki Financial Group) starting April 1st. This initiative supports the DX of screening operations.
**1. Background**
In housing loan screening, scoring models play a crucial role in understanding applicants' credit risk based on application information, credit information, etc., and utilizing this for screening decisions and setting interest rates and guarantee fees. In recent years, with rapidly changing economic conditions such as rising interest rates and inflation, as well as the accumulation of data and the development of AI technology, there has been a growing need for new models that are more accurate and flexibly operable than conventional models.
MRI has consistently supported screening decisions and risk management in the retail loan sector for regional financial institutions for over 20 years through the construction of scoring models and post-introduction monitoring. Additionally, through its "Screening AI Service," which automates decisions on loan applications, MRI has been working on automating and streamlining screening operations using AI. Based on these achievements, MRI has now developed an advanced "AI Scoring Model" for housing loans, which is an enhancement of conventional scoring models, and began providing it for practical use to Mebuki Financial Group starting in April. Mebuki Financial Group is the first to introduce MRI's "AI Scoring Model."
**2. Overview of the AI Scoring Model**
**(1) Features**
**① Improved prediction accuracy of default probability**
MRI's "AI Scoring Model" is an AI model that predicts the likelihood of inability to repay (default probability) and quantifies credit risk based on housing loan applicants' application information and credit information. Compared to conventional models that typically make judgments based on 5 to 10 items, this model can capture more complex relationships from a wider variety of items, achieving improved prediction accuracy of default probability. By accurately grasping credit risk, it becomes possible to make judgments based on objective evaluations even in cases where lending decisions were difficult under previous standards.
**② Design based on the characteristics of each financial institution**
The AI Scoring Model provided to Mebuki Financial Group is characterized by its construction and adjustment based on the group's unique housing loan screening practices and portfolio characteristics, utilizing the group's loan performance data. In addition, based on MRI's extensive knowledge gained from long-term support for regional financial institutions in the retail loan sector, its AI design is strong not only in prediction accuracy but also in ease of confirming judgment reasons and consideration for continuous operation.
**③ Ease of introduction and operation through API linkage**
Furthermore, the evaluation results from this model are provided by linking via API (※2) with the financial institution's personal loan business support system (※1). Since model judgment results can be linked between systems, it can be applied to practical operations without requiring major modifications to the personal loan business support system. As a result, it can be introduced in a realistic time and cost, and after introduction, the model can be periodically and quickly verified and updated in response to changes in the external environment and portfolio.
**(2) Relationship with the Screening AI Service**
**1. Background**
In housing loan screening, scoring models play a crucial role in understanding applicants' credit risk based on application information, credit information, etc., and utilizing this for screening decisions and setting interest rates and guarantee fees. In recent years, with rapidly changing economic conditions such as rising interest rates and inflation, as well as the accumulation of data and the development of AI technology, there has been a growing need for new models that are more accurate and flexibly operable than conventional models.
MRI has consistently supported screening decisions and risk management in the retail loan sector for regional financial institutions for over 20 years through the construction of scoring models and post-introduction monitoring. Additionally, through its "Screening AI Service," which automates decisions on loan applications, MRI has been working on automating and streamlining screening operations using AI. Based on these achievements, MRI has now developed an advanced "AI Scoring Model" for housing loans, which is an enhancement of conventional scoring models, and began providing it for practical use to Mebuki Financial Group starting in April. Mebuki Financial Group is the first to introduce MRI's "AI Scoring Model."
**2. Overview of the AI Scoring Model**
**(1) Features**
**① Improved prediction accuracy of default probability**
MRI's "AI Scoring Model" is an AI model that predicts the likelihood of inability to repay (default probability) and quantifies credit risk based on housing loan applicants' application information and credit information. Compared to conventional models that typically make judgments based on 5 to 10 items, this model can capture more complex relationships from a wider variety of items, achieving improved prediction accuracy of default probability. By accurately grasping credit risk, it becomes possible to make judgments based on objective evaluations even in cases where lending decisions were difficult under previous standards.
**② Design based on the characteristics of each financial institution**
The AI Scoring Model provided to Mebuki Financial Group is characterized by its construction and adjustment based on the group's unique housing loan screening practices and portfolio characteristics, utilizing the group's loan performance data. In addition, based on MRI's extensive knowledge gained from long-term support for regional financial institutions in the retail loan sector, its AI design is strong not only in prediction accuracy but also in ease of confirming judgment reasons and consideration for continuous operation.
**③ Ease of introduction and operation through API linkage**
Furthermore, the evaluation results from this model are provided by linking via API (※2) with the financial institution's personal loan business support system (※1). Since model judgment results can be linked between systems, it can be applied to practical operations without requiring major modifications to the personal loan business support system. As a result, it can be introduced in a realistic time and cost, and after introduction, the model can be periodically and quickly verified and updated in response to changes in the external environment and portfolio.
**(2) Relationship with the Screening AI Service**