Streamlining Legal Operations with Generative AI: Lessons from the GMO Financial Gate Legal Department
Hello, I am Hiroyoshi Teruyama, Representative Director of Corporate GPT Inc.
"AI and DX in the legal department are difficult."
I often hear this sentiment. It is true that compared to sales or development departments, corporate departments—especially legal departments—are often said to lag behind in operational efficiency, AI adoption, and digital transformation (DX). However, I would like to introduce the initiatives of the GMO Financial Gate, Inc. (hereinafter GMOFG) Legal Department, which has taken on the challenge of systematically utilizing generative AI and achieved remarkable results.
The Future of Legal Affairs Opened by Generative AI - Challenges and Possibilities
Traditional legal work involves many complex tasks requiring highly specialized knowledge, such as contract review, legal research, and compliance management, leaving staff constantly pressed by heavy workloads and time constraints. Moreover, each task requires careful judgment. On the other hand, because of the specialized nature of the work, processes often become personalized, relying heavily on the experience and knowledge of individual staff members. This leads to issues such as inconsistent work quality and difficulties in handing over tasks when staff changes occur.
Heavy workloads and chronic time shortages
Inconsistent quality due to the personalization of tasks
Difficulty in knowledge transfer
The Challenge of the GMOFG Legal Department: Details of Generative AI Implementation - Success in Three Stages
To solve these issues, the GMOFG Legal Department adopted a highly strategic approach to implementing generative AI. This approach consists of the following three stages.
1. Visualization and standardization of the entire contract review process
"What should we have generative AI do?"
They first posed this critical question and, to clarify it, began by visualizing the entire contract review process and standardizing the thought process. Specifically:
Breaking down the contract review process into 74 steps
Defining the work content of each step in detail
Establishing standard judgment criteria for each step
This is what they did.
2. Requirement definition for each step and analysis of AI utilization potential
Next, for each visualized step, they defined requirements (work content and clearance levels) by asking, "What does this mean, and what must be done to proceed to the next step?" In this process, they achieved:
Unification of work understanding among members
Breaking away from personalized work
Identification of potential for generative AI utilization
At this stage, regardless of generative AI utilization, it can be said that the contract review work itself within the legal organization reached a considerable level as an organizational function.
(3) Consideration of Prompts for Generative AI Application
Finally, for the processes where the use of generative AI was deemed effective, we designed specific prompts (instructions for the generative AI). The key points here were:
Creating specific instructions based on the requirements of each process
Clarifying the expected output
Setting standards for quality checks and verification
This clarified why using a specific prompt for a particular process is correct, and established the criteria for checking the deliverables produced by the generative AI.
By steadily advancing through these three stages, the GMOFG Legal Department was able to achieve concrete results toward the organizational use of generative AI.
The Secret to Successful Generative AI Implementation: Business Visualization and Requirement Definition - Concrete Results and the Importance of Visualization
As shown, to successfully implement generative AI, it is crucial not only to understand the technical aspects but also to thoroughly visualize business processes and define requirements. In the case of the GMOFG Legal Department, this point was particularly emphasized, and thorough business visualization and requirement definition were conducted through training.
As a result, the following three concrete outcomes were achieved.
Improvement in Output Quality
Rework and backtracking by less experienced staff were reduced, resulting in at least a 20% improvement in operational efficiency per man-hour.Reduction in Operational Time
The total time required for contract review across the entire organization was reduced by at least 20%.Long-term Vision and Strategy
By simultaneously visualizing the contract review process and utilizing generative AI, literacy regarding the risks associated with generative AI implementation (legal risks, data privacy, etc.) was strengthened organizationally, establishing a system that balances operational efficiency and quality improvement while ensuring thorough risk avoidance.
Diagram of Business Process Visualization (Example)
Toward Company-wide Generative AI Implementation: Learning from GMOFG's Initiatives - Challenges and Solutions
The GMOFG Legal Department quickly focused on the potential of generative AI and implemented it with the goal of improving operational efficiency and organizational capabilities. As a result,
① Reduction in operational time for contract review tasks
② Improvement in organizational work quality
③ Improvement of AI literacy across the entire organization and construction of a knowledge management foundation
and other remarkable results have been achieved. This case study is a concrete example of how generative AI can bring innovative changes to legal operations and serves as an important reference for many companies.
Outlook: From the Corporate Department to the Entire Company
We have introduced the initiatives of the GMOFG Legal Department. Moving forward, the expansion of generative AI utilization to areas outside the Legal Department is also being considered. In doing so, the challenges that the company must address for company-wide implementation are as follows:
Clarification of Policy
: Clarify the company's internal usage policy for generative AI.Security
: Strict access control and data protection.Internal Rule Formulation
: Formulate usage guidelines that include compliance (legal adherence).Data Governance
: Thorough data quality management and privacy protection, and the formulation of data policies.AI Education for Personnel
: Improving expertise through the implementation of continuous training programs.Monitoring
: Improving operational rules based on regular monitoring and effectiveness measurement.Information Gathering and Implementation Consideration
: Rapid information gathering on the rapid technological progress of generative AI and consideration of business implementation.
Conclusion: Transformation is Possible
Returning to the opening statement, "AI and DX in the legal department are difficult," the case of the GMOFG Legal Department proves that this stereotype is not necessarily correct. Certainly, careful preparation and a strategic approach are required. However, if these are carried out appropriately, AI implementation can definitely be achieved in the legal department and throughout the entire corporate division.
Generative AI is expected to continue to evolve and contribute to the efficiency and sophistication of operations across the entire corporate division, starting with the legal department. Using the GMOFG Legal Department's case as a reference, the following are important points to consider when introducing generative AI to the corporate division:
Thorough visualization of operations
A phased approach
Organization-wide initiatives
In this process, it is possible to adapt risk management, which is the original work of the corporate division, to the AI era and shift to tasks with higher business value.
*This article examines the possibilities of business reform through the introduction of generative AI, based on the initiatives of the GMOFG Legal Department. When introducing generative AI, appropriate consideration according to the characteristics and situation of each organization is necessary.