Efforts for establishing governance to support AI utilization, in collaboration with Resona Holdings
While the use of AI technologies such as Generative AI is accelerating, many companies face challenges in securing internal controls and complying with various guidelines issued by the government and industry groups. To achieve effective AI governance, it is necessary to respond to the situation of AI usage in the field and the maturity of the organization. NTT DATA offers an "AI Governance Consulting Service" to help companies utilize AI safely and securely. This comprehensive service includes system design, rule formulation, educational support, and technical implementation, such as red teaming and introducing guardrail tools.
As an example of such efforts, this article introduces a practical example of AI governance construction co-created with Resona Holdings (hereafter referred to as Resona HD).
1. Overview of AI Governance Consulting Services
As AI technology advances and Generative AI becomes more widespread, it is important not only to respond to the technology itself, but also to establish utilization policies and control methods for the entire organization to maximize its value. The purpose and business processes of AI use vary from organization to organization, and the way it is introduced and operated varies depending on organizational culture and structure. In addition, risk tolerance differs greatly from industry to industry, so there are many cases in which generic guidelines provided by the government or industry associations do not lead to effective operation in the field.
With this in mind, NTT DATA provides AI Governance Consulting Services with the aim of establishing effective governance in the field. This service supports the establishment of governance that is tailored to each organization's situation, taking into account factors such as organizational structure, maturity of use cases, and usage patterns. It covers everything from system design, such as rule formulation, to technical approaches, such as the introduction of red teaming and guardrail tools, and provides consistent support from policy consideration to implementation of countermeasures.
This article introduces the development of guidelines, the establishment of a risk check process, and the development of educational content for employees as an example of efforts with Resona HD.
2. Initiatives with Resona Holdings
Resona HD is leveraging advanced technologies, including Generative AI, to enhance customer service and transform its operations. The two companies have been collaborating on this initiative based on the shared belief that identifying the risks involved in advance and establishing a system to appropriately control them when using AI is necessary. As part of this effort, we have been working to build governance and improve the effectiveness of Resona HD while exchanging opinions based on its organizational and business characteristics.
2.1 Formulation of guidelines that penetrate the organization
Responding to AI risks requires that guidelines be referenced in practice and that they become a common language within the organization. From this perspective, we worked to formulate effective guidelines.
In formulating the guidelines, we checked for consistency with existing company regulations, as well as for wording and expressions actually used in business operations. The guidelines were designed for practical use, with consideration given to an easy-to-reference structure and an appropriate amount of information. In addition, we collected information on technical and social trends surrounding AI governance as needed and reflected it in the content.
Through this process, we clarified AI-specific considerations in a way that is easy for practitioners to understand and respond to, while maintaining consistency with Resona HD's existing governance framework.
2.2 Establishment of a risk checking process
As the number of projects utilizing AI and data increases, variation in risk assessment from case to case is a growing concern. To address this issue, we have developed a risk checking process that can be applied to all projects.
This process was designed based on a risk-based approach that takes into account NTT DATA's accumulated project knowledge as well as guidelines and regulatory trends of governments and industry associations. In order to facilitate use in actual operations, we have organized the viewpoints required for risk checks and structured the items for easy decision making.
Furthermore, through a step-by-step trial introduction, we worked to promote understanding and establish the process in the field. We are continuously improving the process so that it is easier to use for those in charge, reflecting feedback from the field obtained through the trials.
Through these efforts, we are unifying internal decision-making standards, promoting standardization of governance, and establishing a process that will enable stable risk response in the future use of AI.
2.3 Designing training according to the goals
We have developed educational content aimed at acquiring basic knowledge so that employees can increase their sensitivity to AI risk and respond appropriately in their work. This is positioned as a step in advance of institutional responses to AI governance, and is intended to serve as an entry point for employees who are new to AI risk.
The target audience includes not only AI developers but also general employees who may use AI in the future. The content is written in plain language and can be read in a short amount of time so that it is easy for anyone to understand. In addition, to make it easier to recall specific risks, we have included specific examples based on actual work.
Through these educational measures, it is expected that the risk awareness of the entire organization will be fostered, while employees will understand the background of the guidelines and rules, and will be able to make decisions and respond autonomously. Education plays an important role in enhancing the effectiveness of governance in line with actual conditions, not just in formalities.
3. Future Outlook
The activities in this case, such as the formulation of guidelines, the establishment of a risk check process, and the development of educational content, were all advanced in stages based on the customer's business characteristics and goals. Through these efforts, we are steadily improving the level of governance while ensuring effectiveness in the field.
The framework for AI governance has not yet matured, and it will be required to respond flexibly to changes in technology and social trends. In particular, the effective functioning of governance is extremely important as AI utilization in the field progresses. In particular, AI governance is becoming more important in supporting changes in technologies closely related to business operations, such as Generative AI.
NTT DATA has supported the advancement of AI governance for customers in a wide range of industries and business types, with its broad support experience from the introduction of AI technology to its actual operation and its strong practical knowledge from the field. Through its AI governance consulting services, NTT DATA will continue to contribute to the realization of safe and secure AI utilization by providing a realistic and sustainable approach tailored to the customer's situation and goals.
Endorsement
NTT DATA has provided us with a wide range of support from the initial stage of the AI promotion system, not only in the development of AI governance, but also in both offensive and defensive aspects, including support for the creation of use cases and upgrading of the AI infrastructure. Their support is based on more than just the latest knowledge about AI governance; it is also based on a deep understanding of banking practices and the Resona Group's systems. This has led to great reassurance and results. We look forward to working together as strategic partners to further expand the use of Generative AI in the future.
Takahisa Usui, General Manager, CFT (IT Reform) , Group Strategy Department, Resona Holdings, Inc.
Taro Kato, Group Leader, IT Planning Department, Resona Holdings, Inc.
Mr. Takahisa Usui
General Manager, CFT (IT Reform) , Group Strategy Division, Resona Holdings, Inc.
Joined Asahi Bank, Ltd. (currently Resona Bank) in 1996.
For many years, he was engaged in system infrastructure planning, etc. in the IT department. After serving as General Manager of the IT Planning Department of Resona Holdings, he has been in his current position since April 2023.
Mr. Taro Kato
Chief Manager, IT Planning Division, Resona Holdings, Inc.
Joined Resona Bank, Ltd., in 2012.
After working in branch sales, he joined the IT department. Worked on system development at an affiliated company and internal intranet planning at the IT department. Current position since April 2024.
Honoka Sato
NTT Data Group Technology Innovation Headquarters AI Technology Department
She supports a wide range of customers in utilizing their data, from technical support to human resources and organizational development. After working as a PMO on a large-scale AI application project, she is now engaged in developing AI governance and promoting data management, utilizing her knowledge of AI risk management and data governance.
Toshiyuki Fujikubo
NTT DATA Technology Consulting Division
Digital Technology Director specializes in planning and promoting DX strategies for business transformation and designing and managing CoE organizations. His strength is the combination of expertise in a wide range of IT fields cultivated through the development and operation of multi-tenant banking systems and business consulting.