FY 2026 Digital Agency Information Systems Procurement Reform Study Meeting (1st)

Last Updated:: Oct 20, 2026

Outline

For an overview of the Study Group, please refer to the 2026 Digital Agency Information Systems Acquisition Reform Study Group .

Event Information

Date and time: Friday, October 2, 2026, from 17:00 to 18:30
Location: Online (Microsoft Teams)
Committee members present: Chairman Shoji, Committee members Ishikawa, Kawasawa, Sakashita and Shiroyama

Proceedings

1. Opening
2. Proceedings
- 1. Explanation on the Advisory Panel
- 2. Evaluation of the validity of measures taken in the past fiscal year
  - 2.1 "Improvement of Agile and Flexible Procurement Procedures"
  - 2.2 Outline of Initiatives to "Improve Capability of Outsourcers in System Procurement" and Explanation of Key Issues
  - 2.3 Open Discussion
  - 2.4 Outline of initiatives and explanation of issues for "Expansion of entry opportunities for SMEs, start-up companies, etc."
  - 2.5 Summary of initiatives to eliminate vender lock-in and explanation of key points
  - 2.6 Outline of efforts and explanation of issues in "Ensuring Transparency through Internal Controls, etc."
  - 2.7 Open Discussion
- 3. Explanation of issues and free discussion on concepts for new information system procurement reform measures, taking into account the recent changes in the environment surrounding information systems
3) Closing

Data

Summary of the proceedings

After the Secretariat explained the background and purpose of the Study Group, the status of measures implemented in the past fiscal year, and the candidates for priority issues of new measures for information system procurement reform, discussions were held, and the main opinions from each committee member are as follows.

1. Validity evaluation of measures taken in the past fiscal year

(1) Overall

Individual measures seem to be appropriately evaluated as progress in means to improve procurement. On the other hand, it is necessary to track long-term changes in terms of whether actual procurement has improved as a result of the implementation of measures, whether an adequate system has been built for users, and whether user satisfaction has improved. Even if quantitative evaluation is difficult, if examples of specific improvements achieved by adopting agile development cannot be confirmed, there is a risk that the efforts will become a mere facade.
Since the degree of difficulty and the level of effectiveness vary, it is necessary to check the progress of the measures in the Digital Agency separately from the progress of their deployment to the Japanese government as a whole. It is also necessary to have a mechanism for verifying whether or not new efforts have led to the achievement of the final policy objective by combining factors such as the degree of satisfaction after the conclusion of the contract, the successful bid rate, the number of bidders, and the performance after the execution of the contract
It is necessary to examine a series of effects, not only the shortening and flexibility of procedures themselves, but also whether or not they have accelerated the development and introduction, reduced the work burden on officials, and finally improved the convenience of citizens and residents. It is necessary to set performance indicators according to the purpose of the project, such as the operation processing time, the quality and stability of the system, and the procedural burden on residents, in addition to the satisfaction level of system users in the government, and to run the PDCA cycle.
Since procurement is a means to achieve policy objectives, and a procurement system is not an end in itself, it is necessary to have a mechanism to link the outputs of a procurement system and the outcomes of the entire policy in order to evaluate not only the utilization of a procurement system but also the extent to which the policy objectives are achieved by the use of the system.
"Although the self-evaluation of measures taken in the past is significant to a certain extent, evaluations that can capture not only the progress but also the essential improvement in the quality of administrative services and procurement by returning to the policy objectives are necessary in the future. In designing future measures, it is necessary to make it possible to grasp appropriate KPIs and to consider how to evaluate them."

(2) Direction of measures (1) "Improvement of agile and flexible procurement procedures"

It should not be an objective in itself to increase the utilization rate of a new procurement method. It is necessary to confirm whether the introduction of a new method has improved the competitiveness of the procurement and the speed of development, and whether it has been continuously used in other projects. It is also necessary to organize the criteria for determining what kind of results should be obtained to judge the method to be effective.

(3) Direction of measures (ii) "Improving capabilities of orderers in system procurement"

In order to establish the capability of the ordering party in the organization, it is necessary to measure the quality of deliverables such as requirements definition documents and the reuse rate of template knowledge. It is necessary to distinguish between the fact that knowledge has been accumulated and the fact that it is used in actual procurement operations, and to evaluate whether or not procurement quality can be maintained by utilizing past knowledge even after personnel changes, and whether or not past failures and rework can be avoided in the next procurement *
Even if the Employer's knowledge and skills are enhanced, the effect of reforms will be limited if the terms of the contract or institutional constraints do not allow for appropriate procurement and flexible consultation with operators. It is necessary to examine the impact of contract terms on the entry and proposals of operators, and to organize the matters into those that can be addressed by the contract model or operational ingenuity, and those that require systemic revision. *
It is necessary to sort out the matters that can be handled only by system changes and the matters that can be handled by capacity support. The definition of capacity differs between the ability to select appropriate operations in the system and the ability to change the system itself when the system becomes an obstacle. The consultation service, Q & A, and support system are mechanisms to support capacity, and they need to be distinguished from the capacity actually possessed by the ordering party.

*Base on prior submitted comments

(4) Direction of measures (iii) "Expanding entry opportunities for SMEs and start-up companies"

It is important not only to increase the number of orders to SMEs and start-up companies, but also to remove unnecessary barriers to entry so that companies with technological capabilities can be appropriately evaluated. It is necessary to confirm not only the increase in the number of participating companies, but also whether a system has been established to appropriately evaluate technological capabilities, and whether specific effects have been produced by hiring companies with technological capabilities
If the objective is to lower entry barriers, the lowering of barriers can be evaluated as an achievement. On the other hand, it is necessary to continue to monitor technical capabilities and performance quality from the perspective of whether or not quality has declined as a result of lowering entry barriers.
The fact that the number of participating companies has increased due to the creation of a competitive environment where digital startups are not at a disadvantage can be regarded as an achievement in itself. On the other hand, it is necessary to check the outcome after the execution of the contract to see whether companies with high technological capabilities have actually entered the market. The stage and indicators for assessing policies to ensure fairness of opportunity and policies to improve the quality of procurement results are different.
Increasing the number of entrants is not the same as improving procurement performance. It is necessary to identify data that show that increased entry has led to an increase in the value of procurement for the government and to make a comprehensive assessment in terms of price, quality and technology. It is also necessary to check the quality of new entrants after receiving orders, delivery dates, continuous business performance, and the adoption of technologies and services that could not be chosen previously.
There is a difference between having an application program and companies actually being able to apply. It is also necessary to reform the system surrounding the procurement system, such as by announcing the annual schedule for government procurement at an early stage and creating an environment in which startup companies can prepare proposals, make sales investments, and plan financing.
The entry of SMEs and start-up companies is a means, and it is necessary to clarify the outcome of what they have contributed. If the purpose is the entry of companies with advanced technical capabilities, the issue is how to evaluate their contributions to technical capabilities and procurement quality.
It is necessary to investigate the reasons why SMEs and start-up companies give up bidding after considering it, including the burden of documents and procedures, payment terms, etc. It is also necessary to check how vendors respond to measures to expand opportunities for SMEs and start-up companies to enter the market and change their behavior.

(5) Direction of measures ④ "Elimination of vender lock-in"

A vender lock-in is a situation in which, for technical reasons, the enterprise has no choice but to continue placing orders for the same system or series of development projects to the same business operator. A decrease in unilateral bidding or an increase in the number of SMEs entering the market is not sufficient to determine whether the technical constraints or continued dependence have been eliminated. Therefore, it is necessary to confirm whether business operators are actually switching when updating the same system.
Even if the same company continues to receive orders, the quality of bidding can be ensured to a certain extent if proposals are reviewed and improved, and specifications are prepared to ensure competitiveness. On the other hand, it is considered that the situation where the same company continues to receive orders with the same specifications and proposals as in the past should be improved.
It is possible that the time that has passed since the introduction of the vender lock-in prevention checklist has resulted in little more than formal confirmation. It is necessary to evaluate whether or not not only the use of the checklist but also dialogue with multiple businesses, market surveys, revision of specifications, and expansion of participation in a transparent manner have led to a decrease in single-party bidding or the elimination of vender lock-in.
Vender lock-in is not an issue of unilateral bidding, but an issue of portability, that is, whether it is possible to switch to another carrier when necessary. Obligations relating to data migration and knowledge transfer should be included in contracts, and an assessment should be made as to whether the ordering party can secure necessary information when switching carriers.
If the data format, API, setting information, system configuration, operation procedure manual, etc. are dependent on a specified business operator, even if switching is possible under the contract, actual switching is difficult. It is necessary to clarify the cost and period required for data migration, restrictions on additional licenses, handover of setting information, etc. at the procurement stage. On the other hand, if uniform requirements are imposed on all systems, the procurement cost and the burden on business operators will increase. Therefore, it is necessary to stage exit requirements according to the importance of the system and the possibility of substitution *
If the knowledge is not inherited by the ordering party, the past specifications will be used as they are, and actual portability will not be ensured. The system for the ordering party to accumulate and inherit knowledge and the exit requirements are related to each other. Therefore, these issues should also be considered along with the capability of the ordering party.
Although the ratio of unilateral bidding has been decreasing, the absolute number of unilateral bidding has not decreased despite the increase in the total number of projects. It is necessary not only to check the ratio for each fiscal year, but also to track individual projects to confirm whether there was an actual switch of carriers.

*Base on prior submitted comments

(6) Direction of measures (v) "Ensuring Transparency through Internal Controls, etc."

When post-contract changes increase due to AI and agile development, issues arise in relation to transparency and fairness in the initial bid. When appropriate decisions are made at each stage, a record should be kept and a system for continuous confirmation should be in place.
Internal controls and securing transparency must be viewed as issues related to all other measures, and the PDCA cycle must be repeated.
It is necessary to clarify what is to be achieved by transparency. Agile and flexible procedures or agile development may change the response of the ordering party depending on the situation. Therefore, it is necessary to carefully organize the objectives to be achieved and the contents of transparency to be secured.

2. Concept of new information system procurement reform measures

(1) Composition and priority of priority issues

Among the priority issues, the issues of acquisition and contract models, competitiveness and market formation, and knowledge circulation between the public and private sectors are positioned as an extension of the measures discussed in the past. On the other hand, digital resiliency is another axis issue, and generative AI is a cross-cutting issue. Regarding acquisition and contract models, etc., it is necessary to clarify what is incomplete and what to do next as an extension of existing measures, and to prioritize and differentiate between issues.
It is necessary to sort out new issues such as generative AI or digital resiliency and issues related to the extension of existing measures, and to clarify the strategies for advancing all seven review meetings.
There is nothing strange about the five key points. On the other hand, if the purpose of procurement reform is to improve the value of administrative services, it is possible to set total costs, service quality, recovery from failures, and the possibility of switching to other companies as cross-sectional common evaluation axes
It is necessary to clarify the scope of digital resiliency and clarify the differences between AI and conventional procurement premised on fixed specifications and deliverables.

(2) Procurement and AI Governance for generative AI

As for AI, at least, it is necessary to classify the AI into the following three categories: (I) When the public administration, which is the ordering party, uses AI to improve business efficiency; (ii) When the contractor creates deliverables using AI, and the ordering party verifies the deliverables and procures governance; and (iii) When the system or service itself including ICT is to be procured. function
While generative AI is an unavoidable issue, it is difficult to determine the details at the time of contracting when procuring AI. Therefore, it is necessary to consider this issue in conjunction with reform of the procurement-contract model.
In the case of generative AI, there is a difference in logic between the case where the Japanese government offers favorable conditions based on the existing market and technology, and the case where the Japanese government participates in R & D and the formation of markets. It is necessary to distinguish the cases where mature systems on the market are purchased, where the Japanese government participates in R & D and the formation of markets, and where the Japanese government conducts trials and then develops the market, and to consider these cases in light of the Japanese government's AI strategies.
The use of generative AI should not be a goal in itself. It is necessary to factor in what kind of value is improved by procurement. Since AI governance is finally a problem of how to ensure quality and responsibility, it is necessary to embody the relationship among values, quality, and responsibility.
A flexible approach that allows requirements and technology to change after the contract has been signed can also lead to repeated specification changes and increased costs. In addition to flexibility, contracts need to be designed to manage change and take responsibility for outcomes and costs.
When procuring AI, the problem is how to evaluate and procure items whose capabilities or outputs change as a result of learning. In addition, there may be issues that differ from conventional AI that is based on fixed specifications and deliverables, such as the case where a procurement agent is involved in the procurement.
Generative AI, etc. When procuring a technology whose application and potential have not yet been determined, it is not enough to procure it at a low price. The perspective of domestic technology, start-up companies, and market formation is also important. The method of evaluating the effects of procurements of general information systems and systems or services that include AI may differ

(3) Procurement Ensuring Digital Resilience

With regard to digital resiliency, it is necessary to confirm whether it covers only natural disasters or includes cybersecurity as well.

Digital Agency added that digital resiliency is currently being considered not for cybersecurity but mainly for early recovery in the event of system outages caused by natural disasters, etc.

(4) Future considerations

As for the direction of new measures, a wide range of issues on GIS procurements were presented, such as changes in technology environments including generative AI, methods and units of procurements in view of AI transformation, and agile and flexible procurements procedures. As some issues could not be sufficiently organized in the first round, it is necessary to specify each issue in the next and subsequent rounds based on the opinions and additional analysis this time.

Greater Than