Digital Extraordinary Administrative Advisory Committee Working Group Technology-based regulatory reform Promotion Committee (2nd)

Overview

Date and Time: Wednesday, November 16, 2022 (2022) from 10:00 to 12:00
Location: Online
Agenda:
1. Opening
2. Proceedings
  1. Direction and issues of future discussions of the Technology-based regulatory reform Promotion Committee (explained by the Secretariat)
  2. Explanation from Ezaki Member
    - Proposal for a technology map
      - How to proceed with risk assessments of security, etc. in technology utilization -
  3. Explanation from Mr. Shimada
    - Toshiba's Concept of AI Quality Assurance
  4. Exchange of opinions
3. Adjournment

Materials

Minutes, etc.

Date and Time

Wednesday, November 16, 2022 (2022) from 10:00 a.m. to 12:00 p.m.

Location

Held online

Members present

Chairman

ESAKI Hiroshi (Digital Agency Senior Expert (Architecture))

Members

OKADA Yūsaku (Professor, Department of Management Engineering, Faculty of Science and Engineering, Keio University)
Keiko Ogawa (Certified Public Accountant, Banking Capital Markets Leader LegTech Leader Partner, EY Strategy and Consulting Co., Ltd.)
Shinpei Kato (Associate Professor, Graduate School of Information Science and Engineering, The University
KAWAHARA Yoshihiro (Professor, Graduate School of Engineering, The University of Tokyo)
Yumi Kawabata
Yutaka Saito (Executive Director of the Information-Technology Promotion Agency Digital Architecture and Design Center)
Taro Shimada (Representative Executive Officer, President and CEO, Toshiba Corporation)
Takao Someya (Professor, Graduate School of Engineering, The University of Tokyo)
Keisuke Toyoda (Specially Appointed Professor, Institute of Industrial Science)
Takao Nakagaki (Professor, Faculty of Creative Science and Engineering, School of Science and Engineering, Waseda University)
NAKAMURA Osamu (Professor, Faculty of Environment and Information Science, Keio University)
Ayumu Nagai (President and Representative Executive Officer, Astameze Co., Ltd.
Katsunori Nemoto (Counselor of the Japan Business Federation)
Daiyu Nobori (Director of the Cyber Technology Laboratory, Information-Technology Promotion Agency)
Yutaka Matsuo (Professor, Graduate School of Engineering, The University of Tokyo)

Overview

Secretariat (Suga): . It's time to open the second "Technology-based regulatory reform Promotion Committee".

Again, members are invited to participate online.

Mr. Endo, Mr. Ogino, Mr. Kyuma and Mr. Suzuki are absent due to personal reasons. Mr. Kawahara and Mr. Someya are expected to leave the office in the middle of the meeting.

In the course of today's proceedings, we will have an explanation in the first half and an exchange of opinions in the second half. During that time, we would like to use chat to listen to your opinions and questions. We would like you to use the Slack of the Committee as a chat medium as much as possible, but if it is difficult, we would like you to post your comments in the chat section of WebEx. We will check the comments you posted at the Secretariat, and we would like to pick them up in a timely manner.

Without further ado, I would like to begin today's agenda.

I would like to ask Chairman Ezaki to proceed with the proceedings from now on. Thank you very much.

Chairman Ezaki: Is there anyone else, including .
Let's get down to business.

For the second session, we plan to discuss the direction and issues of future discussions at the Technology-based regulatory reform Promotion Committee, the second, how to proceed with the evaluation of risks such as Technology Map in the use of technology for security maintenance, and the third, AI quality assurance considered by TOSHIBA.

At the end of the meeting, I would like to hear the opinions of all the members. I would like to have a chat with you, and Mr. Someya and Mr. Kawahara will come first, so I would like to have your comments as appropriate.

First of all, Mr. Hitoshi Suga, Director of the Secretariat, would like to explain the direction of future discussions and points of contention in the Technology-based regulatory reform Promotion Committee.

Secretariat (Suga): materials, and this is a review, but it explains the outline of the efforts of the Digital Consultation. Page 3 is also a review and summarizes the goals. There are no particular changes.

The next page is the table of contents, and this time, I would like to explain in accordance with the three pillars.

First, we have conducted a preliminary public offering, and based on the learning from that, we have compiled a proposal from the Secretariat on what kind of information should be posted in Technology Map and how to think about it.

Second, based on the lessons learned from the preliminary public offering, we have prepared a draft proposal on what kind of target area and target should be considered first, and what kind of process should be followed to compile the map.

Third, I would like to explain some of the remaining issues, such as the editorial policies, the design of incentives that can be used, and the Trust of collateral for the content of the article.

First, the first pillar. From the next page is the ideal way of listing information, and on the next page, we have summarized the outline of the public offering that was implemented in advance. Thankfully, although it was a short period of time, we received 24 applications from various people, from large companies to small and medium-sized enterprises and venture companies. Currently, according to several criteria, after detailed examination by the secretariat, 20 applications are published on the website of the Digi-cho.

On the next page, as items to be included in the catalog, we held in-depth conversations with the regulation authorities in advance, heard that they are concerned about these matters, and included them as essential response items. The yellow highlights are a summary of what the authorities said they would like us to hear. The next pages are all items to be included in the catalog, but mainly we are concerned about whether we can take thorough measures against fraudulent entrance examinations in national certification.

The next page summarizes the learning.

First, as a matter of course, cooperation with the regulation authorities is essential for this catalog, and through dialogue with the ministries and agencies responsible for regulation in In-person Courses, we have learned that the detection of fraud, such as spoofs and cheating, is essential, and we have asked for answers.

In fact, some business operators have pointed out that if some essential functions are not provided, even if other functions are enhanced, they will appear to be inferior. At present, however, we believe that this is unavoidable in light of the needs of procurement operators.

We also took the time to show the results of the public offering to the ministries and agencies responsible for the regulation and the organizations conducting the tests and lectures, and asked them if they could use the results. Many of the comments were very positive. Some said that they would like to see the outline of the technology at a glance and not to write it down freely. Others said that they would like to see the evaluation criteria and technical level clarified. Others said that they would like to see a layout that is easy to compare and search because it is published in a simple format and cannot be searched. Others said that they would like to see some ingenuity so that business operators do not have to access each device individually because they are worried about the introduction cost.

In any case, we reaffirmed that it is necessary to cooperate with the authorities even before the public offering is implemented.

The second major issue is that regarding the public offering this time, the secretariat considered the training and examination to be one process, and the public offering was made for products that can cover the entire process from application for the training to payment of the fee, implementation of the training, implementation of the examination, and issuance of the certificate. However, many companies said that they wanted to apply, but they declined because it is impossible to cover all the cases with full specifications.

On the other hand, there are some element technologies, from large to small, but in extreme cases, there were cases where companies that only have identity authentication technology or companies that have technology that can detect human movement with a camera did not meet the initial requirements for public offering set by the Secretariat, but they applied by filling in the necessary items column appropriately. At present, these element technologies are posted on the website as a separate frame because you have applied, but I once again thought that it is necessary to decide how to handle them in advance.

There are various possibilities, such as utilizing the creativity and ingenuity of start-up companies, discovering new technologies that we are not aware of, and having companies cooperate with each other in elemental technologies. Considering that the unit of technology search is a convenience that the Secretariat has decided on its own, at this point, I believe that it should be allowed to apply for only some of the elemental technologies. I would like to hear your opinions.

Third, if we say that this second issue is OK for a wide range of elemental technologies, another problem will arise. In the future, in the case of identity authentication technology, I think that we will be asked to apply again and again as an elemental technology to all public appeals for catalog maintenance. Regarding common technologies that are expected to be utilized across multiple technical fields, rather than accepting them as elemental technologies in individual fields many times, it may be better to extract them separately and maintain a catalog. I think this is also an issue.

In addition, I have written three other points.

Going forward, through dialogue with the ministries and agencies responsible for regulation, first of all, if you think that the technology you applied for can be adopted as it is, I would like you to proceed to procurement. Before that, in regulation, if there is a regulation that cannot be done without a face-to-face meeting, it will be reviewed. On the other hand, if it is determined that this application information alone is insufficient and a validation is necessary, a Technology validation will be implemented, and I believe that the necessity of a Technology validation will be determined through dialogue with the ministries and agencies responsible for regulation.

Second, I said that there were some cases that were not posted after detailed examination by the secretariat, but one of them is services before release. It is very helpful that there is a need to apply for marketing because it will be an advertisement, but in fact, the secretariat does not have the capacity to follow whether it will be released for sure, and fact check and validation are difficult at this point, so this time, I automatically dropped it from the target of publication.

In addition, there are actually several applications after the public offering period has ended, and although I think it is not very fair, we have made additional announcements this time for convenience.

On the next page, I have organized the advantages and disadvantages of posting main listing information, including the above.

The first is whether or not the essential requirements considered by the authorities, which I mentioned earlier, will be made essential, which I explained earlier. Also, elemental technologies will be included, but how about preparing a separate catalog for cross-cutting technologies?

This is a new point, but we have received many comments that the best practices are actually useful. By publishing the best practices, which are adoption cases, the context in which they are applied and operated, as well as the cases, can be firmly shared in addition to the technology itself. It is very useful that they can judge that they can use it because they want to use it in a similar way.

On the other hand, some have pointed out that as a disadvantage, it may result in the exclusion of start-ups that have not been adopted, and that even if they have achieved results, there may be cases in which the market should be opened first through a review of regulation because there is regulation and the market itself does not exist. Although we have not yet been able to grasp specific examples of problems, I believe that the current assumption is that Technology Map will include best practices, but it will not be impossible to include them without essential requirements and results.

In the next column, I would like to see the disadvantages and restrictions of provision included. It is natural to point out that even if only what can be done with technology is explained gorgeously, information on what cannot be done is more useful from the user's point of view.

On the other hand, as a disadvantage, there was a comment that even if we communicate with the companies that applied this time, to be honest, the companies that have the technology can provide information on restrictions that should be noted, but it is difficult to write down the disadvantages or things that cannot be done because the companies that have the technology have the right to decide whether or not to apply in the first place, so they will not apply if it will be negative marketing. Therefore, regarding the necessity of listing, I think it would be good if there was a mechanism in which the companies would express the points to be noted, and on the other hand, the users, including the ministries and agencies responsible for regulation, would supplement the information that they have used.

Then, as I mentioned earlier, it is a matter of price. The side that wants to adopt it says that price information is essential for on-site introduction decisions, and it is impossible if it is too expensive.

On the other hand, when I talked with companies about this, it was also pointed out that the selection focusing only on price was too much ahead of the government, so I would like them to evaluate the performance and order the performance. In addition, there is naturally a case where the price is set flexibly after understanding the requirements and scale of the user, and it is quite difficult to set the price uniformly. Therefore, although it will be optional for the time being, as you pointed out in the Consultation Session, I think it is possible to consider cooperating with the digital marketplace when it is established.

Also, in terms of other matters, as I stated below, I would like to not forget your point that Cyber security should include information on what kind of certification it has obtained, what kind of standards it conforms to, when the information was obtained, and information on the related guidelines.

On the next two pages, I would like to hear your frank opinions on the current ideas. I would like to know whether it is good to continue doing business a little in the future, rather than just inviting applications once.

First, there are many regions where inspections, inspections, and diagnostics are conducted on-site or on a regular basis. This regulation is the subject of the seven preceding reforms, and there is a need to know the technologies necessary for this.

The scope of inspections, inspections, and diagnostics is extremely wide, and there are needs for diagnosing the risks of accidents, theft, and fires in land, sea, and air areas and sections, as well as for monitoring the state of specific structures and natural objects, diagnosing degradation risks and disaster risks, predicting them, and determining the timing of repairs by estimating the remaining life. Regarding the second point, in particular, the existing catalog created by the Ministry of Land, Infrastructure, Transport and Tourism is very well made, and cooperation between them is also important, and it is a little difficult to handle even if you suddenly search in such a wide area, for example, 2,000 or 3,000 applications, so we would like to consider what scope of technology search should be conducted first, including such areas.

In terms of target technologies and search technologies, sensor monitoring technologies are represented here. In addition, I believe image analysis will also be covered. Since there are all kinds of sensor technologies, I have described the items in the catalog below as images. Since they are extremely complex options, I would appreciate if you could give me advice on how to do it.

Second, I would like to change the color a little, but a privacy tech company has a technology called secret calculation, which is actually a technology to protect privacy. When this technology is used, secret calculation processing can be performed without transferring the personal data and causing the person who is directly transferred to touch it. Such a technology has been developed, but according to the interpretation of the current personal data Protection Act, no matter how secret it is, it is still a personal data. However, as an operational ingenuity, it is interpreted that highly encrypted data is not regarded as a problem. However, it is unclear whether this highly encrypted data is applicable to the use of secret calculation, so there is a problem that the implementation of technology has not advanced. I would like to introduce this as a possible target.

In addition to this, we have received consultations on several interesting themes, so we would like to continue to do so.

The second pillar is from the next page. This is the state shown in the figure. When there is a vertical regulation or no Technology Map, and when there is a horizontal technology or no technology, I would like to sort out where the region or catalog should be targeted.

First, on the vertical axis, if there is no regulation, it is left blank at the bottom, but from the purpose of the Digital Consultation Initiative, I believe that this will be temporarily excluded from the target. If there is a regulations on paper and in-person processes, there are cases where it is acceptable to use the technology, and cases where it is unacceptable to use the technology, and if it is unacceptable, I believe that the regulation will be reviewed as much as possible and the restrictions on the adoption of technology will be lifted.

As for the horizontal axis, of course, we would like to focus on the case where there is technology, but it is ambiguous between the case where there is technology and the case where there is not. There are quite a few zones between the case where there is technology and the case where there is not, such as the case where it is not certain unless the performance is confirmed or the case where the performance is not achieved. We believe that it is important to set the scope of technology search by taking as much as possible those that are said to have no technology into Hirome so that changes in the middle area are not overlooked. Therefore, the entire area is shown by the hollow dotted line.

Within that range, if there are definitely mature technologies and there are no regulation restrictions on the use of technologies, there is a possibility that we can suddenly go to the brown area of Technology Map listing. If we list all of them, it may collapse, but if there is a need, we can list them on the map even if there are no regulation restrictions. If there are restrictions, we will list them as if there are no restrictions when the regulation is reviewed.

In terms of technology, I think it is difficult to make a completely objective judgment on the distinction between areas where mature technology is definitely available and areas where we want to confirm the performance or where we have not achieved the target, but we will make a judgment in some form, and if it is judged that technology validation is necessary, we will first ask them to do so, review the regulation, and then implement the process of listing it on the map. In addition, if there is a regulation, and we would like to innovate compliance using technology, but it seems that there is no technology yet, we will write development target on the far right side. I have organized it conceptually so that the authorities can set such a target.

On the next page, based on that, I have considered four steps as a process for maintaining Technology Map.

The first is to categorize regulation as we are doing in the Digital Policy Consultation. We will extract regulations on paper and in-person processes and categorize them based on the purpose of the regulation being implemented across ministries and agencies. We will discuss what the authorities are concerned about in a certain category, what the essential requirements for regulation are, and what the risk tolerance is. It will be an extremely difficult conversation, but I believe we have no choice but to clearly state these to the extent possible. The items in the catalog that I have been talking about will be clearly stated, and the process in which such requirements are reflected in the public solicitation and the items in the catalog will be necessary first.

Next, we will enter the technology search phase in the form of soliciting technologies that contribute to the review of the classified regulation. While looking at the technologies that came out at that time, we will have to have a dialogue with the authorities again to determine whether or not a technology validation is necessary. At this point, I think the primary criteria are whether the authorities are worried about it and whether they think it is impossible to say that it is good without a validation. If you have any other wisdom, I would like to hear it.

On the other hand, for mature technologies that are considered to be unnecessary for validation, we will suddenly publish them in the map and catalog after the minimum fact check.

If the authorities say that a validation is necessary, we will go to the third process and implement a technical validation, and if the performance is insufficient or the technology does not exist, we will set a development target.

After the Technology validation is completed, the final process will be to review regulation. I believe that the purpose of this initiative is not to review regulation but to connect the technologies listed in the map and catalog to the specific procurement by the public and private sectors.

On the next page, in order to categorize regulation, when we generalize the purpose of various regulation as an abstract concept, what kind of technology is there, and what is widely used at present, what is under technology validation, and what has the possibility of utilization but has not yet been unlocked, are tentatively placed in the table to the extent that the Secretariat can understand. Actually, I think it is interesting because it contains a lot of information when read carefully, but I think that unless we do this work first, the phase of regulation categorization and technology search cannot be completed.

On the next page, the point is cooperation with the ministries and agencies responsible for the regulation. First of all, in the development of the Technology Map, we are currently conducting validation in order to build a mechanism in which the Ministry of Posts and Telecommunications will pay for the validation of the Technology budget request across ministries and agencies.

On the next page, the ministries and agencies in charge in regulation have pointed out various matters that need to be coordinated, and I believe it will be necessary to clarify the essential requirements and what I have just described in the process diagram.

The next page is the third pillar. As for the editorial policy, the summary of the discussions so far summarized by Chairman Ezaki is arranged by item.

First of all, I think it was about how to think about the main body of editing, issuing, maintaining, and operating the catalog. Then, it was about the responsible body and the scope of responsibility when the technology listed in the catalog was used. I think this may be the most serious. At present, the catalog is provided only as reference information, so the editor of this catalog is not responsible for anything. I think we will consider whether it is really good or not.

In addition, regarding the target of providing the catalog, the first target will be the decision maker of the government agency in charge of regulation. However, we would like to go to procurement after that. Therefore, we need to increase the number of articles that appeal to people who make decisions on organization within the government's regulation or the organization of companies subject to procurement, and those who have substantial technical decision-making authority as pointed out by Mr. Noboru last time.

In addition, I believe that the functions that we would like Technology Map to have are to match needs with seeds, which will expand our dreams, to have texts with technical support that would enable us to understand how technology will change the future, and to have functions that automatically generate maps based on the axes and ranges that users of the catalog desire.

In addition, I believe that the text editing method must be in a technology-neutral format that is easy to process automatically and use for secondary purposes.

Continuing on to the next page, it is an incentive design for continuous update operation. Initially, of course, the Digi-cho will be launched by the government, but I think you pointed out that it can be developed as an ecosystem led by the private sector, rather than standing in as a gatekeeper. In that case, what kind of organization and capability is required for maintenance and operation is a matter to be examined.

In addition, as Mr. Noboru pointed out, we will thoroughly consider a system that makes it easier for technology holders to renew their knowledge and a system that stimulates the motivation of people to write about it.

In addition, we would like you to use the map in addition to creating it. In terms of incentives for use, it is necessary for the ministries and agencies responsible for regulation to establish a mechanism to review Technology Map by referring to regulation. We are currently considering this.

In addition, for technology holders, we will consider cooperation with DMP, which I introduced earlier, so that technology that is directly procurement by the government will be connected to government procurement as smoothly as possible. As an incentive for business operators, I believe that the value of this catalog will increase by firmly setting essential items to be observed in regulation, and by properly cooperating with the ministries and agencies responsible for validation when determining the necessity of technology regulation, and by firmly drawing out their thoughts. At this point, I am writing so.

It has been a long time, but on the last page, I wrote that it is the mechanism of Trust collateral, and it is like this now.

The first is that in the preliminary public offering, the Technology Catalog is only reference information and is not certified or approved by the government. In addition to that, I think we are working to make it meaningful by cooperating with the ministries and agencies responsible for regulation. In addition, I am writing that it is possible to make the status of acquisition of various certifications, including security certification, appealing, and to determine whether it will be reflected in the procurement requirements in the future.

If there is any other good system, I would like to hear your opinion.
That's all from the office.

Chairman Ezaki: Is there anyone else, including .

I would also like to ask if you have any questions or comments on the explanation from the Secretariat, as there are some members who have already attended the meeting or who are attending the meeting in a very difficult situation, and I believe that many of them were organized as I mentioned earlier.

, please.

SHIMADA Member: I think it was good that the materials were very well organized.

My only question is about incentives. The actual phenomenon is that, for example, from the perspective of gas companies, even if they reduce costs by introducing IoT, the corresponding profits are suppressed and do not benefit them. If we do such a thing, there is a Issue in which it is meaningful for the company to take risks and make efforts in the first place, so I think it would be good if such a design is also considered.

That's all.

Chairman Ezaki: Is there anyone else, including .

That's exactly right. If we don't make profits, people don't invest, so I would like to share our practices on the incentive design part.

Mr. Nakamura, please.

NAKAMURA Member: I think it was good that various things were put together.

One point is that it is a feedback on the part that needs to change the regulation and the law, but I feel that it is reversed. The purpose of the regulation, what kind of regulation is there for what purpose, and the breakdown of what quality is necessary to achieve the purpose have not been made. The method of simply testing something three times a year or doing it once every five years, and doing a written examination or how many hours of training should be done, is probably written in the law. I feel that this is the essence of the reform.

Now, when making the map, I think you are probably talking with the ministries and agencies that are conducting regulation and discussing what the purpose is. I think that can be done in the same way for all regulation. In other words, we will make it a little clearer what the purpose is and what we have to do. I think the method has been developed first.

How many tests should be done, how much testing should be done properly, for example, what should be inspected once a year with a sensor, and how much visual inspection should be done. It is very important to break down the purpose of what is being inspected and what is being done in order to write a map, and it is important to break down the purpose of what is being inspected in order to satisfy the purpose of what is being inspected, so I feel that it will not go around unless it becomes a story about what kind of methods are available. I feel that this is probably the part where the Digi-cho will be waving its flag.

Chairman Ezaki: Is there anyone else, including .
Mr. Suga is also nodding.

Secretariat (Suga): When we assumed that regulation was rational until now, the wave of the digital age has come, so let's review whether regulation is really rational even if it is digital once again. This is what we are talking about with each regulation authority. I think it is a process that is so costly that it is daunting, but I also think it is a key point. For example, speed restrictions are applied to ensure that cars can run safely without contact with people, but I think it is to reach a conversation that speed restrictions were the best means even if automated driving technology was assumed, so I would like to do it sincerely.

NAKAMURA Member: , perhaps creating a catalog is one method, and in the communication about creating a catalog by that method, I feel that it would be good if we could ask the regulation Ministry and Agency to reconsider such a part once again. I feel that it would be good if we could support the work of so-called breakdown.

Secretariat (Suga): .

Chairman Ezaki: Is there anyone else, including Maybe, it would be good if cases reviewed in such a way were posted as practices. I will guide everyone to do it by referring to this.

, do you have any other opinions?

Then, I will make time for you at the end, so if you have any opinions, I would like to ask you to organize your thoughts or statements at the end.

Next, I would like to briefly explain how to proceed with the risk evaluation of Technology Map and others in the utilization of technology for the development of security, along with Material 2.

What is important is the current practice, which will change over time, as Mr. Suga said earlier. We must not forget to consider how the private and public sectors will support each other, and it is not just a problem for the public sector.

In addition, since it is meaningless unless it is introduced and operated, I believe that how it was introduced and operated is not only a problem of the technology map, so-called technology, but also important. As I believe it is related to Dr. Nakamura's talk earlier about sharing this practice, it is not a matter of having such a method, but how to have a method of introducing operation, such as reviewing it, and migration, which are very important practices to share.

As for incentives, as President Shimada said, it is necessary not only to provide incentives, but also to share practices on how to respond to cases where digitalization costs increase or do not lead to cost reductions, and if that is possible, to reflect it in policy. In that case, it would be better to include in this map how to share knowledge, knowledge, and experience across industries.

The third point is that we have to consider not to be hampered by the excessive demands and regulation that you mentioned.

In the next page, I will talk about the obvious, and it is essential to avoid such a thing, so we must proceed. In that case, I think we must properly anticipate the essential requirements such as Cyber security, which Mr. Suga mentioned.

In that case, as I mentioned earlier, I am writing one more thing, including rephrasing it, but in terms of what current practices to share, I believe that not only knowledge but also experience in operation will be very important practices to share. As knowledge, it will be easier to study and write, but what should be the method of operation, and of course, introduction and operation will be discussed, and in the case of incentives, I believe that the key will be how to design incentives, including money.

In addition, as validation is mentioned, it is extremely costly for the government to do this, so it is probably necessary to provide information that will lead to how to industrialize this, and I think it will be very effective to introduce cases where this has worked well. In that case, I think that the government will support and the private sector will lead the creation of an agile validation system.

In addition, basically, it is easy to talk about ASIS as a map, but I think we need to think together about changing it to To Be, that is, to change it to a new function or a new structure.

Then, as a way to make it, from KPI to KGI, from the materials of today's Secretariat, it is written that we need to change what is decided only by price as a problem awareness. Then, I think it is necessary to make KGI a requirement of procurement as a practice.

In that case, the Cloud by Default is the one that the government was able to successfully drive to be able to do these four things in one.

I participated in the discussion on Cloud by Default, but it is not just about sharing data. In the second place, by leaving it to experts, the reduction of expenses as labor costs or the development and securement of human resources will be reduced. In addition, CAPEX and OPEX will be reduced, and CSR and BCP will be included in the fourth and fifth places. If we are not satisfied with this, Cloud by Default will not be effective, so I think we will make this a requirement. I think it is necessary to properly share KGI instead of KPI as a practice.

Also, if we consider that digitalization will take this means, manual work, which is in the first person, to digital work, which is in the third person, basically, conjecture and falsification of data will not be possible with digitalization and online. Auditing will also be very simple. Whether or not this will be true is, as Shimada-san said, if there is no cost merit, private sector will have no incentive to enter. What do you think about this?

In addition, considering that responsibility lies not with people but with computers, when Web3 comes in, psychological stability will come out to employees as a utility of digitalization and as a KPI. I think it would be good if we could include this in a good way.

On the last page, as one of the important requirements of the technology map, I think that the security requirement should be made so that information can be provided and understood in one stop. Systematically speaking, it is important to evaluate not only the so-called cyber security, but also the operation stability design, so that the equivalent will not come out for the place where it was done by people.

Operation is also important, and in addition to simple elemental technologies and system requirements, it is necessary to share methods for how to operate well and to share incentives for doing so well.

With regard to the validation System of Technology, I believe that it is also important to consider how the private sector can industrialize the validation business, and to guide the people so that they can promote industrialization by presenting appropriate examples.

Also, in relation to the issue of price alone, while changing KPI to KGI, new methods and incentives, such as Minister in charge of Administrative Reform Okada or carbon neutrality, are considered as one of the axes, but not limited to this, how to increase incentives and how to bring them to places other than price should be covered as a technology map, so this is my proposal.

That's all. Thank you very much.

Next, Mr. Shimada, could you explain Toshiba's concept of AI quality assurance?

SHIMADA Member: Then, before I say about Toshiba's concept of AI quality assurance, I would like to say that we need to carefully consider the operation that you have pointed out. I have prepared this. I am going to say what is the process to ensure that the operation can be done properly. I think you can understand it if you look at the content.

After all, automation of inspection and other such things means that the quality is guaranteed. It must be possible to confirm whether the required performance is satisfied. However, quality assurance is not easy for AI. I believe that AI should be treated in the same manner as humans rather than considered as robots. In other words, AI is useful, but the problem is that stable quality cannot be achieved unless it is properly trained.

Specifically, as I believe all of you here are well aware, it is impossible to achieve 100% prediction accuracy. It is extremely difficult. In addition, we do not know why the results were obtained. In addition, it is difficult to guarantee the operation after the start of operation. As I will explain later, it is often said that the system deteriorates over time, but if the accuracy decreases or the system is used in an unexpected way, the results cannot be predicted at all.

So, AI quality assurance is a good tool, but we need to ensure the black-box is transparent. In addition, we need to consider the quality of the training data and the fairness of removing bias. In addition, we need to design the AI model by paying attention to the fact that it needs to be protected by false recognition, fail safe, and so on. This is very important for the quality of the training data and the quality of the AI model, and we need to look at the transparency, fairness, accuracy, and robustness of the data to be trained. Even so, the answer may not be correct. What is very important is to guarantee the quality of the system. I was originally engaged in aircraft design, but the people who are engaged in aircraft design do not trust computers. Therefore, if we want to make it fly-by-wire, we need to make it quadruple. We must consider that the same thing will happen to AI.

For example, when considering the automated driving of a car, we are currently trying to tackle this with AI in various ways, but for example, when AI produces strange output, it is possible to think that a person can override it or that it is corrected by automatic control.

Various measures are taken in AI. For example, if AI is also duplexed, it is difficult for me, as an aircraft engineer, to say whether it is okay to use duplexing, but I don't know which is wrong. I think it must be more than three, but we can solve this problem by using a transparent AI in combination with a rule-based AI that always gives the same answer. In addition, by creating a monitoring system that protects against abnormal output, we can improve the robustness of the AI system with high image recognition in automated driving, even if it is extremely stable and output is manually output to the driver.

Using this kind of thing, for example, in the maintenance of plants, in the case of AI for Anomaly Prediction and Detection, in the first place, in the case of plants like this, there are many cases where indent products and indents are dedicated products, so it is a dedicated AI, but there are many cases where the performance of AI is not improving. Therefore, by feedback of AI data while the maintenance staff judges the necessity of maintenance, etc., and by adding more data, AI and humans gradually work together to improve the function. In fact, we provide this kind of thing.

This is another example, so I would like to make a brief comment. Toshiba Carrier is also using the mechanism I just mentioned for early confirmation of refrigerant leakage. This is also being implemented in the water treatment and sedimentation tank automatic monitoring services of Kurita Industries.

For manual quality assurance, quality is determined by the performance of the AI system to meet customer expectations. Learning data, AI models, control systems, and operations are important elements of AI system performance, and AI quality assurance is to maintain the performance of the AI system to meet customer expectations. In other words, it is necessary to guarantee the quality of learning data, AI data, control systems, and operations as a whole. Even so, the quality of AI is not sufficient, and quality assurance as a system needs to be done in combination with other things.

To this end, we have created guidelines for AI quality assurance. First, we have established guidelines for quality assurance, and then we have decided on a quality assurance process to ensure the quality of the model. Based on this, we will guarantee the quality of the model, and when the customer operates the model, we will provide a quality card so that it will not be a problem. We will propose that AI can be used as a whole.

The guideline is to obtain basic information on what everyone should be careful about as a whole. It is a basic information such as a viewpoint list, what viewpoints are there, and what should be cared about. As I said earlier, AI is like a human being, so unfortunately, it is better to think that AI does not work as it is once it is introduced. There may be some AI, but when it is used for surveillance or other purposes, since it is a human being, if it is a robot, we tend to complain about why it does not work. However, it is natural for humans to come in after school and not be able to work properly, and we must consider the education system as well.

We have created a process for this. This is a general process, so if we do this according to the process, we can stably operate it. This is an example, but for example, what is the Issue of the business, is the obtained data OK, is uncertainties properly considered, is re-learning possible, and so on. When actually operating it, we always think about the quality of the data, the model quality card is in the customer's hands, and we must carefully watch it. This is an example, so please take a look at it later.

Using these, we will introduce and operate a system that uses AI to analyze drones images from power transmission line inspections, as well as image diagnosis of products in the manufacturing process. I think you can see the contents of this later.

Finally, MLOps is a continuous update of machine learning. Basically, AI is often said to be rotten, but it sometimes degrades. We must understand that AI often degrades in accuracy or gives completely different answers due to changes in the characteristics of the equipment. Therefore, from the development phase to operation and maintenance, we must carefully consider operation maintenance that can feedback this information and become smarter and smarter like a human during use. Maintenance of AI models involves retraining of AI models, so we must consider a mechanism that can continuously do this.

So, what I wanted to say is that we have to think about the operation as well. As Mr. Nakamura said earlier, it is absolutely true, but if we cannot define a process to do this and many people cannot understand it, then in the end, it is true, but we cannot implement it.

I would like to explain a little about the Appendix. In fact, I think that Japanese people are leading the world in terms of AI quality assurance. I carefully watch Japanese trends and the creation of rules in the international community, but I feel that the rules emerging from Europe and the United States concentrate on the perspective of how to protect individuals from risk and how to operate them appropriately by applying regulation.

Among the research institutes of Japanese academic societies, there is a group called QA4AI. It is an AI product quality assurance consortium, and it seems that people who really like AI get together on their own. On the other hand, what is important is that more than half of the people involved are quality assurance people, not AI people. In fact, Japan is an advanced country in terms of quality assurance, and I think it is important that many people who have knowledge of various quality assurance systems and mechanisms have been involved in the production of world-class Japanese products. For a certain period of time, the whole world has studied this.

By incorporating this into the development of AI, I believe that we can promote the application of digital technology to the world, not only AI.

In addition, to be blunt, I believe that it would be better for the Government of Japan to issue guidelines to study the process before regulation, and I believe that it would be ideal for the Government of Japan to make its first transition in regulation when the guidelines are completed and specific details are made available.

I would like to make a brief presentation. Thank you very much.

Chairman Ezaki: Is there anyone else, including , thank you very much.

In terms of operation, you focused on AI and quality assurance. Processes are also not simple technologies. They are extremely important, so how to share them is necessary. As an example, I think you talked about AI, which is very closely related to digitalization.

Finally, the fourth point is that we have about one hour to exchange opinions. I would like to hear the opinions of those who will leave due to their time. If you raise your hand with the show of hands function, I would like to ask that person to speak.

Mr. Someya, please.

Someya Member: . I would like to ask two questions about Dr. Ezaki's explanation.

In the fourth point of the final summary slide, you mentioned carbon neutrality and Minister in charge of Administrative Reform Okada. You explained carbon neutrality, and I understand it very well. I think it is an extremely important point from what perspective Minister in charge of Administrative Reform Okada needs to be considered in KGI. I would like to ask you to explain this.

Another point is that there is a possibility that the cost will increase in digitalization, so I believe that your point that incentives are extremely important is true. Under such circumstances, for example, in the case that digitalization and analog products coexist in a transitional manner, it is impossible to avoid the cost increase, so I would like to hear your thoughts on how to overcome the situation and add incentives. Thank you.

Chairman Ezaki: Is there anyone else, including Then, when it comes to my Minister in charge of Administrative Reform Okada, I think that the supply chain check is one of the important points that you have already recognized. From the perspective of service sustainability, it is extremely important to determine the relationship not only in terms of technology but also in terms of operation, including the structure of risk management's business. However, there are situations that vary considerably depending on external conditions, so we should continuously evaluate where there are risks and what we need to think about.

In addition, I believe there will probably be a broad direction on where to focus from the perspective of the Minister in charge of Administrative Reform Okada. In terms of how to formulate a view on this, in a sense, I believe that we must be extremely careful, while also being aware that the Government of Japan will not interfere too much.

In terms of incentives, I don't think there is one way, so it will be about how many successful cases can be shared. In that case, for example, support funds and research development support as policies will contribute to accurate and effective measures while sharing the fact that there are cases where they worked well and cases where they unfortunately did not work. Also, as a business, there will be many companies who have successfully overcome difficult problems by using such methods. I think the point is how to share this.

Someya Member: I understand very well. Thank you very much.

Chairman Ezaki: Is there anyone else, including ?
On the chat side, Mr. Kato said that naming is very important.

Member: , I think it's a waste of time because there is a completely different thing called a technology roadmap. I think you are talking about regulation and certification, so if I say Technology Map, I don't see it. I think I know more about it, but I think I know more about regulation and certification, and I think it's a national story. I think I know more about it, but at least I see it. My name represents my body, so I think it's worth considering at this timing.

In terms of technology, I think regulation is included this time, so if I say technology in general, especially in Japan, there are many real technology companies and real hacker companies, so in a sense, I think it will be quite misleading. Therefore, I think authentication and regulation are very important this time, so if you don't make it clear, the target users won't see it. At least I think I'm one of the target users of this, but if the government releases a press release saying Technology Map has been released, I think quite objectively, I don't see it. I think it would be better to make it easier to understand, like regulation or certification, and remind them of it.

Also, as Noboru-san mentioned, I think the opposite is also true.

From a corporate perspective, I think this is already a policy. When it comes to technology, Accenture and McKinsey have great reports that you can look up by yourself. But in regulation, there are such things as the Ministry of Economy, Trade and Industry, the Ministry of Foreign Affairs and Trade, and the National Police Agency. If they are organized properly, I think they will be a very valuable map or catalog. That was my first naming story.

Chairman Ezaki: Is there anyone else, including There is also a slight nuance of the Chinese character regulation and regulation.
Suga san, you react to this.

Secretariat (Suga): Yes. When we decided the name of the committee, I think we had an essential discussion about what we would do in the first place. It is the same for Technology Map, and now the name does not have any simple edge. I think it is an important essential issue, such as whether we will include the keyword regulation or whether we will have a function of certification. I have written a little in the Secretariat's materials. Even if we cannot go to certification, I think we must first discuss how to secure Trust and how to take responsibility.

In addition, as Mr. Noboru pointed out, we would like to make it possible for people who are not interested in regulation to refer to it. On the other hand, what I wrote a little in the Secretariat materials this time is that we conducted a preliminary public offering and reconfirmed that there is a large part of added value in visualization and verbalizing the ideas of the regulation authorities. I believe that there are aspects that cannot be separated from this work, such as visualization, rationalization, abstraction, and categorization of the ideas of the regulation authorities, which are at the top of the assumed users of the map. Therefore, I think it is also a matter of consideration whether or not to keep regulation away.

Chairman Ezaki: Is there anyone else, including is. It is not a simple explanation of the rules of regulation. What should we do?

Mr. Nakamura, please.

NAKAMURA Member: As expected, this is a position story, and Professor Ezaki also uses the term "private sector-led." For example, there are various projects in private sector that are being publicly invited, such as a project to conduct some kind of test. If there are such successful cases, will they be put on this map, and if there is such a thing that can be used, will the government use it? Or on the contrary, will the name change depending on whether the so-called government is positioned as a map to take the lead in changing the world by using such technology?

Simply saying a technology map this time sounds like saying that there are various technologies in the world, please tell me about them, and the government will make a digitalization by referring to them. To me, it does not mean that I am reluctant to do so. If the so-called government takes the lead in digitalization and makes a catalog that shares successful examples with everyone, I think it will look different from the outside. When I talked about whether or not the elemental technology mentioned earlier will be included in the catalog, if it is written that the elemental technology can be used in this way, it is very interesting, and I think that many people will put it on the catalog and work with the government.

However, if we simply think of it as a document for the regulation authorities to refer to, saying that they can do this kind of thing in the world, and that they have such a track record, then we will do it, it will be less interesting, or rather, it will not have much meaning. I wonder if it will simply become a site that collects and summarizes information. I hope that it will be a catalog in which the Digi-cho is led by the government, and I hope that it will not be led by the private sector, but will be a combination of the government and the private sector, and will be able to publish more and more information on the digitalization of technology related to government regulation.

Chairman Ezaki: Is there anyone else, including .

Your opinion is that it is necessary to be fully aware of the fact that there are several ways out, and that we must be able to respond to demand. Amidst this, the Government must take the lead in showing that we can change.

Nemoto-san, please.

Nemoto Member: , but I think it is important to clarify the purpose of recruitment on the procurement side, and I think it is necessary to make a catalog of what kind of technology can be used to ensure.

On page 11 of the material that Mr. Suga first explained, there was an expression such as a description of disadvantages and restrictions. In fact, I don't understand this story well. The administrative agency should have made the necessary announcement, and it is sufficient that it can be achieved. I feel that it is too much to ask for everything. That is one point.

In addition, there are problems in various areas in the Points to Note. What will be done about the introduction of technology into areas where there are business monopolies? There are a considerable number of business monopolies, and they are also carried out as a form of regulation, so new technology can be used in those areas. However, I think it is the basic stance of the current administrative agency that only people with XXX qualifications can be treated. In that case, it will be a form of declaring from the beginning that technology will not be introduced into areas where people with qualifications are working, so I would like Digital Agency to work hard to a considerable extent on this part.

Furthermore, as mentioned in the previous discussion, there is a tendency to require a 100% guarantee only when a machine is used. As mentioned earlier, there was an example of whether or not there was a function to prevent falling asleep when conducting a test. Even if a human is a proctor, it does not mean that falling asleep can be prevented 100%. I felt that the catalog would not be completed unless the level of recognition was properly included in the requirement.

That's all.

Chairman Ezaki: Is there anyone else, including , do you have a response?

Secretariat (Suga): If you say so, it is true that you need disadvantage information in addition to the essential requirements that we have set up through dialogue with the authorities. I thought it was necessary to take a closer look at what disadvantages were talked about. The authorities will first say that what is considered essential is essential. On the other hand, as Mr. Toyota said in the chat earlier, I actually want to discuss the possibility that it is not essential in the first place, and I thought that was the first concern. Assuming that the essential requirements are sufficiently wide, I think it is certainly important to consider whether there is a need for further disadvantage information in addition to that.

Also, regarding the monopolization of business, where it is written that there is no regulation on page 15, is it correct to understand that you have mentioned that there are cases in which there is actually a regulation?

Nemoto Member: In short, if there is a gray area answer that there is a possibility of illegality like Legal Tech, it means that the administrative agency is saying that it should not be done by machines. It means that we are afraid that the logic will be that it is not good to replace with machines in all fields where business monopolies are being carried out.

Secretariat (Suga): The case that Mr. Nemoto just mentioned is in the upper left of this figure. In other words, it may be necessary to confirm whether there is technology, but at least in terms of regulation, there is a regulations on paper and in-person processes where only humans can do it and only qualified people can do it, and it is in the category where the restrictions on the use of technology are extremely severe, so I think we will consider it as a target.

Nemoto Member: Nice to meet you.

Chairman Ezaki: Is there anyone else, including If you don't organize the terminology well, the meaning of business monopoly differs from person to person, so you have to do it carefully. And the disadvantage is that it may be written as a point of attention or something like that. It may be used in a very bad direction.

, please.

SHIMADA Member: Right now, I just wanted to make a presentation because I wanted to say that there are some things to keep in mind rather than disadvantages. I think it's something like, you shouldn't think of AI as a magic wand, AI is a cute child like a human, so you can't use it effectively unless you make a proper education.

Secretariat (Suga): Shimada, regarding the disadvantages, in the discussion that we should prepare a disclaimer at the same time, such as a rule that users must read this before using it, or a disclaimer that they may not be able to do this, or that they should use it thinking that there is a possibility of an error, we may write that they should read it properly as a point of attention for the technology.

SHIMADA Member: That's right. I think most new digital technologies in the future will use AI in some form, so I have to get rid of the expectation that I can use it as it is after buying it.

Also, I wanted to tell you about the process when we apply it. It is a guideline, we study it properly, we have an operation card at the time of operation, and we start to guarantee the quality as a whole, which I think all of you here know, of course. By carefully promoting such things, it will be possible for ordinary people to use it for the first time. That is my argument.

Chairman Ezaki: Is there anyone else, including .

Regarding the guidelines and guidebooks for operations and such places, Mr. Kato wrote that the guidelines and guidebooks should be used properly.

Then, Mr. Kawahara will be here until eleven twenty five. Thank you.

Kawahara Member: : I myself have been involved in various DX, and the most problematic case is that there are rules written like this, so when I think that I should just code this as it is, there are actually a lot of implementation processing that seem to be judged comprehensively in the end, and even if I try to follow the rules written, it does not go so easily. So, it can be automated, but at the end, it can coexist with AI, which tends to decide on the parts that need to be fine-tuned. Rather, I felt that the best practice I really want to know is how to handle flexible operations cleanly in the system. This is my first point.

The other point is about creating incentives. As Dr. Someya commented, there are quite a few disadvantages in the short term, but in the long term, there are advantages if everyone works hard to share the cost. I think this is exactly the case with vaccine certification. It can be used for overseas travel or for other purposes, and the government is confident that it will be better if everyone shares the cost. I think it is creating incentives to create rules with strong leadership.

That's all from me.

Chairman Ezaki: Is there anyone else, including .

Incentives are a really difficult issue, and this is where the administrative officer's ability will be shown, but I hope that such information collection can be done well here.

On the chat side, we should make sure to follow the guidelines for reading guidebooks, and make sure to maintain guidebooks, including operation, and write down incentives.

Member: I think the naming is important because when I saw this, I said that I should stop using it because I don't know which is better, the guidebook or the guideline. I think ordinary people don't know the difference. The guideline, the guidebook, and the logic are such things.

Chairman Ezaki: Is there anyone else, including If we organize that, it is very important to display the web pages and outputs created here by clearly distinguishing them by category.

Member: I feel that if all Japanese people could understand the difference between guidelines and guidebooks, Japan would become a very strong country.

Chairman Ezaki: Is there anyone else, including , do you have any other opinions?

Mr. Nakagaki, please.

Nakagaki Member: Waseda University. I'm currently working on smart security with NITE, and I have some concerns.

Earlier, I believe that the integration of AI operations is exactly the same, but one major issue is who will own the data of competitive areas and non-Tokyo competitive areas. For example, when it comes to security and maintenance, it is a certain asset that is targeted. If the owner of the target asset has a public institution, such as civil engineering or electrical equipment, it is fine, but if a certain company owns it, it is the data owned by that company.

What is more difficult is that it is the business operator who provides the maintenance service. It is often the case that the company that delivered the plant takes over the maintenance service. In that case, we want to lock in the data and use it to secure our own superiority. When that happens, disclosing or sharing the data will eventually benefit others, so I think we want to ensure confidentiality. We don't have money and we don't have time, so if we want to save labor as much as possible, I think the purpose is to raise the level of the whole. On the other hand, I think we should also secure the profitability of the company.

For example, wind power is a particularly good example. For example, when overseas manufacturers try to collect data for condition-based maintenance, they secure all the data, and even the owner of the asset cannot access all the data. There are cases where the data does not spread in such a state. Therefore, we have such a problem.

Rather than raising a problem, I made a statement about Issue, which I am currently facing.

That's all.

Chairman Ezaki: Is there anyone else, including .

That may be what I am talking about, but the fact that it is such a problem is shared, and in other industries, there are probably some that have solved it by this method. For example, in the computer industry, it is said that all the logs are taken by users, and in the medical care industry, the rules were revised so that individuals can take all the information that hospitals had at first. If this happens in the global domain, the problem is how to create a global consensus through cooperation between the supervisory ministries and industries.

Then, as the output of Digi-in, the output is to do it properly, and it comes up from here. In other words, one of the purposes of this committee is to escalate the fact that there are tasks that must be exceeded in the consideration of actually advancing digitalization, and to put them into policy. This is probably what Mr. Nakagaki said.

What do you think, Mr. Suga? He may not want to do such a thing.

Secretariat (Suga): Nakagaki to be a member of this committee, I thought it was reasonable for him to point out that point as a concern.

I would like to ask you to make a regulatory reform so that a certain technology can be introduced. I thought that it would be possible to provide guidelines that not only allow the use of technology but also allow the adoption of, for example, real-time monitoring technology, and that monitoring data should be adjusted so that facility owners can access it and the authorities can access it if necessary. I would like to hear your comments and advice on this matter.

Chairman Ezaki: Is there anyone else, including ? Personally, I believe we must do so.

Nakagaki Member: If there are any other cases that go beyond these points, please tell me about them, and I myself am very worried about this area, so I would like to study them.

Chairman Ezaki: Is there anyone else, including It would be good if we could move from the Guidelines to the Guidebook, where there are policies and practices that can be shared and how to solve problems, so we can move forward with specific measures.

In addition, I would like to ask Mr. Nakamura if information such as best practices and operational processes will be included in the map this time. Today, I believe that there are several opinions that it would be better to do this, including opinions from Mr. Shimada and myself.

Mr. Suga, how about this?

Secretariat (Suga): Yes. First of all, as I specifically wrote about best practices in the discussion, I would like to publish it in a way that it does not become a barrier to entry for startups.

In addition, I understand that it is necessary to provide information on the operation process, which was proposed by Mr. Shimada, in a firmly linked manner.

On the other hand, I would like to ask for advice on who should compile the guidebook and how to do it, and what to ask at the stage of public offering.

Chairman Ezaki: Is there anyone else, including , please.

SHIMADA Member: I think it is very important to be an independent organization. In short, I don't think some people should write guidelines or guidebooks for their own convenience. In that sense, I think there is a motivation to incorporate organizations across industries and everyone wants to be able to use them. So, I think it would be good for experts to see them and issue guidelines that the government can accept as guidebooks.

Secretariat (Suga): For example, I think we will continue the process of soliciting applications for technology exploration. We will set up a space for a guidebook to be referred to at that time or some kind of document with a clear operation process. We will first propose it to those who respond to the public solicitation. If more than one document appears, we will assume that it is highly neutral and worth referring to across industries and services, and we will look at it carefully. What do you think about this process?

SHIMADA Member: As I mentioned earlier, since it is an industry organization, something like a QA4AI has been launched, so I think it would be good to use it. If such a public offering is made, rather than an industry-wide organization, individual companies will make various proposals for their own convenience for the purpose of sales. Toshiba is no different after all, but to be honest, I wonder if it will be really good for everyone.

Chairman Ezaki: Is there anyone else, including In that sense, Industrial Cyber security, which I am involved in at METI, is working on a system in which common guidelines for a certain area and industry should be created first, and then a guidebook should be created based on the guidelines for each industry. If it is good enough to create a guidebook, it should be promoted to a guidebook published by the national government. Do you mean that it is good to be able to do something like that?

SHIMADA Member: That's right.

Chairman Ezaki: Is there anyone else, including This may be one of the best practices.
It is very good if such a thing happens in various industries. At that time, there will be another place where we will not forget the global perspective. Now, the story of the Michelin Guide is written in the post.

Mr. Kawabata, would you like to talk about something?

Kawabata Member: After a very high-level discussion here, as an example, the Michelin Guide provides information to users in this way. It provides information that is beneficial to users. It is a guidebook. The guideline, for example, has very easy-to-understand regulations for 1 to 3 stars.

At the same time, there are very strict internal standards for selecting restaurants, and the training of people to create standards is exactly the same as the AI I mentioned earlier, but the Michelin Guide should not be mistaken. For restaurant people, if a restaurant falls below three stars, it is worth dying. Then, they are required to be as strict as a machine, and the training of those people is done strictly by establishing global standards, and it is as if all of them are trained by employees. I think I can imitate that, and I think it is quite good because it has been going on for 100 years.

Another important point is that Michelin, which is the main body of the restaurant, has nothing to do with the restaurant project. When asked about the purpose of making it with what, Michelin says that it is making the guidebook with the aim of having mobility used by safty and security. Why? Because if mobility is used for safty and security, it will drive many distances and go far, and as a result, tires will be sold.

This is what you said earlier that it is difficult to set incentives. The incentive why you are motivated to build such an advanced restaurant or an extremely comfortable hotel is that it will be evaluated by Michelin in the end. It will be published in the guidebook, but I think Michelin is creating a quite correct route, even though it is a very far route, in which it will be good if our tires will eventually sell. There are quite good points in the guidebook where incentive design has been going on for a long time and the guidelines attached to it, so I was wondering if incentive design would proceed smoothly when the discussions so far became realistic.

Chairman Ezaki: Is there anyone else, including .

If you put it together, it may be the document that is most referenced by officials in Kasumigaseki or local government. Incentives design is a very important place, so it may be good if various examples are essentially organized.

Kawabata Member: There are quite a few examples like this, and I cover quite a few systems that are working well in the world, so I think it would be good to think that even though it is not digital, the method of creating the system is actually the same, and if we make it digital, it can be done quickly.

Chairman Ezaki: Is there anyone else, including , please.

SHIMADA Member: I talked about incentives, so to talk about the actual state of incentives, for example, the amount of money paid to the company is already limited, and there are many incentives in the regulation industry, but even if we reduce costs ourselves, it will not be possible to reinvest in our company. I am saying that design is being implemented, for example.

Chairman Ezaki: Is there anyone else, including Of course, there are talks about various incentives, and how we are solving them may be due to our know-how as a company that we cannot disclose, but if there are things that can be shared where possible, if we can refer to them well as successful cases, everyone will be happy.

Ogawa-san, please.

Member: I'm Keiko Ogawa from . Nice to meet you.

I think there were many stories that were very helpful. You talked about the guidelines, but I would like to talk a little about the Trust's collateral framework in the last paper. When considering this framework, I think it is very important to clarify the roles of Digital Agency and each regulation authority from the perspective of roll and responsibility.

For example, as a framework to ensure Trust, I think it is very important to build a governance system that clarifies the roles of each ministry and agency. For example, each ministry and agency should learn from the basic policies for risk identification set by Digital Agency, and identify specific risks. For example, each ministry and agency should set rules for high-level governance policies at Digital Agency. Monitoring validation is also included. In accordance with these high-level rules, each ministry and agency should take responsibility for the control that is clearly defined.

What are the risks in that case? I think we need to sort them out so as not to cause confusion. For example, as Committee Member Shimada mentioned earlier, there are risks from a technical perspective, such as black-boxing of AI or bias data. In addition, there are risks specific to the latest technology itself, such as Issue of RPA and robotics, and Issue of third party risks if cloud computing is used. Even if all the risks are identified and all the risks are grasped and perfect, if the identification of risks that hinder the original purpose is lacking, I think the final purpose of Issue cannot be achieved. regulation

Therefore, how to organize the definition of risk is very important. In addition, we must make decisions on how to respond to the risk. I think it is very difficult to completely eliminate the risk. Responses to fraud are especially so, but we are always in a cat-and-mouse situation with responses. Therefore, in the end, other committee members said earlier that it is very important to clarify rules and processes, and I believe that is also very important for ultimately fulfilling our accountability.

A general risk approach is to comprehensively identify possible risks, analyze the importance of the risks when they are actualized, and depending on the significance of the risks, take the risks that have not been dealt with until now, or reduce, avoid, or transfer the risks by developing controls. I will make sure that I can explain such a risk approach by visualization. In addition, the situation always changes dynamically, so I think it is necessary to have a framework that responds to agile by PDCA.

In particular, the point to be considered this time is that small and medium-sized enterprises and new start-ups will also come in. Then, the problem of cost will also come out. I think there was talk of incentives, but in addition to cost, there will be talk of feasibility, such as whether or not risk can be dealt with in the first place. It will be important to what extent risk is taken in balance with risk and whether or not accountability to the people is fulfilled. Therefore, I think it is possible to prioritize areas where risk is not so high and risk response costs are not so high.

In addition, as I believe you talked about security earlier, there are two types of risks. These are general risks and risks specific to each project. For example, the risks involved in general, such as Minister in charge of Administrative Reform Okada issues including the supply chain, cyber security risks, and information risks, which I mentioned earlier, are different in nature from the risks specific to the project, such as the risk of not being able to cheat in the case of the project conducted ahead of this time. In particular, it is necessary to eliminate the sense of duplication by sorting out the overall risks across the projects.

As the Committee members mentioned earlier, I am aware that various measures are being taken elsewhere. If various measures, such as the Zero Trust Architecture in June, are to be taken, I believe that it is necessary to advance general risks in a manner that does not overlap or fail to be consistent with such discussions.

In addition, one thing that I am not sure about this time is data. I have seen the Standard Guidelines for the Promotion of Digital Government, and it has a data strategy. I believe that the primary purpose of this time is to automate analog administration and digitalization, but on the other hand, in addition to the issue of who owns the collected data, I also think that we cannot ignore the arrangement from the viewpoint of data strategy, such as how the government will utilize it.

In the business world, all of them have started to set up data strategies with DX, but even if they want to do it, there are very many places where there is no data to use in the first place and the affinity is not guaranteed. As you know, all of them are taking on the challenge of AI, but despite the fact that there is a huge shortage of AI human resources in the market, in reality, about 75 to 80 percent of the work of AI human resources is data collection and cleansing. Therefore, I think it is easy to think that the data strategy will be a very heavy work when we proceed with the data strategy later, thinking that it can be done by others, not by this scope. Therefore, I think it is better to think about the generated data, for example, affinity, interoperability, standardization, and which data to target.

Finally, for example, when conducting a validation or an audit, one thing that will cause a significant backlash later is that an audit trail or a digital trail cannot be obtained even though a digitalization has been made. It should be noted that there may be cases in which a validation cannot be made ex post facto, such as the loss of data in an audit trail after a certain period of time, the absence of a time log for a sufficient period of time, or the inability to retrieve data through a search. I believe that it will be important to conduct conditional design, including monitoring at a later time.

Chairman Ezaki: Is there anyone else, including Audit.

There are people who have not yet made statements, but I would like to ask if Mr. Saito and Mr. Matsuo of the IPA would also like to make statements.

Mr. Matsuo: I think you are absolutely right about AI that Mr. Shimada talked about. I think it is very important to firmly grasp these points in order to utilize AI.

My question is about Technology Map. Mr. Suga organized it neatly, and I understood it very well. In order to make Technology Map successful, it is necessary to have the cooperation of many people, and since it has a fairly strong social aspect, it is probably important to consider how to create a system that can be used by many people and how to make it a service. In that sense, I thought it is very important to start using PDCA at an early stage, and to accumulate ideas so that many people can use and participate in Mt. Fuji while conducting various trials.

That's all.

Chairman Ezaki: Is there anyone else, including .

Saito-san of IPA, please.

Saito Member: I have been listening to a lot of things, but I have something to tell you. Basically, this time, the Digital Rincho, which is active in Digital Agency, is talking about creating an ecosystem and creating guidelines and guidebooks there. In a sense, in a platformer-like way, I think it is good to talk about how to revitalize it by providing incentives.

When Mr. Shimada's story came up and we talked about quality assurance, he talked about creating a framework and circulating it properly, including that. For example, when a new technology is applied to a so-called normal regulation, he went into what kind of framework should be used to actually do it, and conversely, he talked about creating it in the committee here.

In that case, in a sense, it has become an image that the ecosystem itself must consider the framework and process of how to handle the current data such as quality assurance. And in Mr. Ogawa's talk earlier, on the contrary, the discussion has developed to the point where we are collecting data here and thinking about how to handle it. So, I have to confirm once again where Digital Agency is thinking. I don't know how far we should consider, so I would like to confirm something. Is that okay?

Chairman Ezaki: Is there anyone else, including , if there are so-called missing areas, there are probably ministries and agencies in charge, and I think the action is to properly work on them. I think Digital Agency will create them if it is under Digital Agency's jurisdiction, but in other areas, the supervising ministries and agencies or the industry will create them properly. So, as I introduced the story of Industrial Cyber security, if such guidelines or guidebooks are created, in the guidebooks, operation and quality assurance are very related to operation, and I think it will be a repository or a technology map prepared by the Digi-cho, or a one stop reference from there, although the name has not been decided yet.

Saito Member: In response to what Mr. Ezaki just said, I would like to add one more thing. In that case, for example, when we actually started operating it, when each ministry and agency saw some guidelines and said they were used, some trouble occurred, some incident occurred, and then, on the contrary, did they come back here and fix the so-called process or framework?

Therefore, in terms of how far the Digi-Agency will go, in a sense, I think it would be good if the Digi-Agency would be responsible for the common parts in the case that the Digi-Agency manages the whole and uses it in the end, and we would be responsible for that. This time, assuming the so-called life cycle after creation, if we had to develop the framework itself in a more proper form, along with the operation and maintenance of the current guidelines and guidebooks, if we were to do that with the Digi-Agency, I think it would be possible to leave the operation organization itself to an external QA4AI. I thought it would be necessary for the Digi-Agency to have ownership while leaving such things to external parties. I thought, how about this?

Chairman Ezaki: Is there anyone else, including , please.

Secretariat (Suga): At this meeting, I believe that you are conducting a constructive discussion on functions that do not yet exist, so first of all, we will firmly discuss that such functions are necessary within the government. We will also discuss that the actors responsible for those functions must bear these responsibilities as a set. I think we can discuss only in the order of who can bear them.

At this point in time, I think it is necessary to consider it responsibly to some extent here, because we cannot just ask for the introduction of technology and leave the rest to others.

Saito Member: I understand.

Chairman Ezaki: Is there anyone else, including should solve very difficult problems properly, and Mr. Noboru, who is related to this in the chat, said that in a structure like organization, where it is very strong, people with a variety of people probably do not move, so if we do not make a structure in which people can move independently a little more, it is dangerous to create a very large frame or a very farm frame.

I think Digital Agency gave me a very difficult assignment at the end.

Nobori-san, please.

Member: As I posted on the chat, I myself had worked on a project at the IPA and the Ministry of Economy, Trade and Industry in the past, and at first it seemed to be a bad situation, but there was an example where it went very well. What is important is that there is a great diversity of valuable members. And everyone, I think it is a great gathering of people who are willing not only to give their expert opinions, but also to move their hands and do implementation on everything that was said.

However, I think that the ideas and fields of expertise will never be the same. If we proceed with it based on a planning principle based on the consensus of all members, it will be the least common multiple of implementation, and there is a risk that the abilities of each person will not be utilized so much and that there will be no social value to be particularly mentioned. In general, I thought that this would be a good idea, but if we continue to have meetings like this for two to three months, and have this interesting meeting, I think that the framework of the greatest common denominator will be naturally determined. If we are one company, I think that we should set a code of conduct to the extent that we can do it, but we cannot do it fatally.

After that, about three months later, and over the next one to two years, I think it would be good for each of us to proceed independently based on our own judgment and at our own responsibility. If we decide on the framework in advance, there is an agreement that it is okay to do most of the work in a comprehensive manner. So, I think it is okay for each of us to do what we want, even if we do not get permission or agreement from each other or make adjustments because we do not agree with each other. This is step (2).

With regard to (4), if you do so, I believe that you will produce abundant results. Although it may be inconsistent because it is not based on a planning principle, I believe that it will be a result that covers a wide range of areas in depth. And since everyone has a natural ability to cooperate, I believe that the results that seem to be inconsistent will be naturally combined and will become natural results as a wonderful digital government that works organically. Based on my experience, I thought that this method would have a higher probability of success than the method of proceeding through consensus formation.

That's all.

Chairman Ezaki: Is there anyone else, including .

In that sense, I believe that some members also said that it is important to move things first, so I believe that we will move things without waiting and make changes while moving them, and in that process, as Mr. Saito said, we will pay attention to what kind of large framework we should create an ecosystem.

It will be difficult for you, Suga san, but let's work together with everyone's cooperation.

Secretariat (Suga): I was a little excited because I thought you had suggested a direction that would make it easier for the Secretariat. Thank you.

Chairman Ezaki: Is there anyone else, including , who has not made a statement?

Secretariat (Suga): Toyoda members via chat.

Mr. Toyoda: In , it may be a little different from the current discussion. It may lead to Meta, but rather than a big implementation, for example, a trial public offering will be made for individual applications or small-scale companies, so I feel that the current discussion and the granularity are not easily connected.

In addition, from the perspective of the supplier, if each company makes a list of what it can produce, it will end up just listing what it can produce now, and new industries and technologies will not be born. Therefore, unless a mechanism for sharing information as a social sublayer, or a mechanism for sharing information not only as information but also as environmental or hardware, is born from here, I feel that real information, re-editing of the industrial layer, and re-value will not be born.

This is also a chicken and an egg, but by doing what is called Technology Map, if you arrange them as a map on a single basin, the granularity is not always the same, but they are arranged separately. So, I thought that it would be necessary to collect them from the perspective of using them to create vertical layers as subcategories, and if possible, to create a two way system in which they can be sent as a request for something like this by first sending them as a forecast.

That's all.

Chairman Ezaki: Is there anyone else, including .

In that sense, in addition to the repository mentioned in the chat, it may be necessary for this committee to define and create a mountain that can be extracted from the repository strategically or empirically. In your opinion, it is a common opinion that a mere repository is not enough.

Secretariat (Suga): Secretariat, I wanted to write on page 17 that we are about to start doing that, but I'm sorry that I couldn't explain it well. I don't want to talk about individual application layers such as lectures and exams, but I have a motivation to structure what kind of function the technology performs when it is summarized and abstracted.

Chairman Ezaki: Is there anyone else, including .

There is still a lot of discussion, and I think it will be a great thing if we do it with alcohol. However, since it is the time that I received, I will conclude today's proceedings.

Finally, I would like to ask about the next explanation of the committee from the secretariat.

Secretariat (Suga): The next committee meeting will be the last one within the year, but it will be held from 10 to 12 o' clock on December 1. The secretariat will contact you again for details.

Regarding today's proceedings, we would like to publish it on our website after the Secretariat has confirmed the draft minutes with everyone who attended the meeting. In addition, if you do not have any particular objection to today's materials, we would like to publish all of them on our website.

I would appreciate it if you could attend such a high-level discussion for a while. Thank you very much for attending the Committee today.

Chairman Ezaki: Is there anyone else, including .

With that being said, I would like to conclude today's meeting. Thank you for your hard work.