Thank you very much for the introduction. Let me introduce the Originator Profile technology we are developing, which is trying to tackle with disinformation and misinformation. I'm Shigeya Suzuki, Professor at Keio University, and also I'm Chair of the Technological Development of the Originator Profile.
So, everybody knows about this context, but let me provide some context we have. So, unverifiable information is dominating the digital landscape as everybody experiences.
So, we are enjoying the Internet, which will provide a communication platform for everybody on the Internet. So, we can send and receive the information freely. That is really a breakthrough, but...
Excuse me, the timer is not clicking. Yeah, please.
So, unfortunately, with the information coming from, we are not sure about where the information is actually coming from, and also about whether the content is true or not. So, it is provided that we are facing a very large risk.
So, this actually... Of course, we are receiving an email, but we are facing this kind of information in multiple contexts. But let me look into the three cases. First is the finance.
Of course, this kind of information can provide a kind of effect to the financial market, of course. And also, as we see in the past several years, the information technology affects so a lot on the election result.
And also, on the disaster situation. Let me discuss about the disaster situation a little bit more.
In Japan, last year, on January 1st, we had a huge earthquake hit in the northern part of Japan. And in that kind of situation, of course, we unfortunately have a good experience on it. And always, misinformation, disinformation is circulating around. And that affects a lot to the people who are working on this sort of situation, like medical professionals need to go somewhere, request it, and can't find somebody who is in trouble, for example.
So, that will waste the amount of human resources in the context, and so on. And also, we know that generative AI and the rise of sophisticated cloud is one of the other problems. We can't distinguish the real picture and the fake picture generated by AI. There's no way to distinguish these two of them.
So, fraudulent actors can make use of this kind of technology to make it look really real to the typical people. And also, if we look at the advertisement market, ads on fraudulent sites escalate the issue. And usually, this advertisement deal is happening real-time on the site, that people who want to show the advertisement will publish the advertisement. Then some company, like a newspaper company, provides a place to show the advertisement. And this is matched among the demand and the supply, and then displayed on the user's client.
But unfortunately, the verification of each of the advertisers, or the publisher who is trying to provide a place for the advertisement, is not verified. So, that causes a mismatch, an unfortunate mismatch between them. And in the worst case, some of the criminals are making use of such kind of mismatch, unverifiable mismatch, to show their advertisements they want.
So, this kind of ad fraud is actually a really big problem. So, we see that 867 million U.S. dollars are damaged annually in Japan. And if we are looking at that globally, it's more than 2 trillion dollars.
So, this has happened because the root cause is a lack of source verification. So, most of the content lacks traceable origin, and others traded without origin confirmation. This is a systematic failure in verifying the content, and also about the stakeholders.
So, this is the reason why we wanted to introduce original profile technology, which enables the user to identify the originator or publisher of the information on the Internet, rather than certifying whether the content is correct or constitutes misinformation. So, this is what we are currently working on.
So, what OP is, and what OP is not. So, OP identifies who sends the content, and confirms content hasn't been tampered with by using the digital signature technique, and does not judge if the information is true or not.
So, if you look at this, then this is why I would like to introduce this to the EIC. Because, actually, we are telling you that this is the original profile, but it is actually a discussion of the digital identity in a certain context. Especially in the context of some of the information provided from, for example, a newspaper company, or a broadcaster, or such kind of company.
So, we are currently targeted at the misinformation, disinformation on web-based media, such as newspapers, companies, and TV stations. And also, we are looking at local government bulletins, for example. As I mentioned, for the large earthquake, or such kind of crisis, then the local government wants to deliver the correct information to the people in the area. And that is very important from our point of view.
And also, we are working on how to address some of the problems happening in the advertisement world. So, what OP provides for the web is two things.
First is, we can secure part of the webpages. It is not necessarily an entire webpage, but a part of the webpages, by the content attestation. And the second thing is the origin identity as originator profile.
So, just a single page of the content, if you are looking at on the news site, it will consist of potentially multiple news sources, and also include some of the advertisements on the same page. Each of them comes from different originators.
So, originator profile can distinguish each of these contents, which comes from where, and also provide detailed information about the origin. So, the structure of the OP is that it consists of the core profile and the profile annotation. Core profile, or CP, is actually a digital signature key material and very essential attribute. And the profile annotation is provided by multiple third parties, the verification of the attributions or self-claims from the originator. These together form a verifiable and tamper-proof rule of OP.
And a single core profile can be attached by multiple profile annotation from multiple annotation providers. So, this information is synthesized by the information verified by multiple organizations.
So, we could say that the profile annotation is a kind of verifiable attribute. And multiple profile annotations from independent annotators provide a way to provide a rich set of verifiable attributes in a decentralized manner.
So, each actor can provide independently, of course, or potentially it can provide multiple profile annotations by one profile annotator. So, from the end user's point of view, they can build multiple verifiable attributes provided by multiple sources.
So, in other words, this horizontal profile technique provides multiple information from multiple sources, not by single authorities. That is the difference of the originator profile technology.
So, this is one of the screenshots of the originator profile. It includes just one source, one article on the Japan Times news site. And this left-hand side is a dimmed part and a highlighted part.
So, the dotted line is marking the area where the contents are distinguished from the inside of this web page. And the upper left-hand side corner, the JT mark, is the icons for the Japan Times.
Then, on the pop-up window in the right-hand side, it provides the origins information. And each of the articles is provided as small icons by the left-hand side.
Then, if multiple contents appear on a single page, there will be multiple contents icons there. Then, clicking on this contents icon will provide the origin information. And you can drill down this information in detail towards each of the profile annotations.
So, we designed this technology in a common, generalized part and an application-specific part. And the common technology part consists of the identity and the content attestation technique. And also includes a baseline governance model. How to govern these profile annotators is an important thing to govern the entire ecosystem.
And also, we implemented web-content-specific technologies, including web-content authenticity provided by the content attestation for web. And also, it provides a presentation as an extension I have shown in the previous slide.
And also, we provide a use-case-specific governance model to design the entire ecosystem. So, we started working on this technology more than three years ago. And we ran experiments in fiscal year 2023 and 2024 with support from the Japanese government. We worked on the news media and also the early experiment on the digital advertisement area.
And also, on the local government information dissemination to address natural disaster situations. And we are also working on an experiment for providing information through news aggregators.
So, it is possible from the news aggregator sites to verify the origin if the news aggregator site is implementing our technology. So, current development status is initial development completed and adjusting features.
So, we start gradually deploying the technology to the actual media outlets. So, we are working together with some of the major news companies in Japan. And we are going to deploy the technology in the calendar year 2025. At this moment, we do not have some flexibility in the identity part, including key rotation.
And also, we want to provide more original profile data placement flexibility. So, we are working on that, but that we can finish in the near future.
So, we also want to implement the actual implementation for the outlet via news aggregators. And also, on the digital advertisement.
So, we are working on it this year. And it will be feature complete for the static website by the end of this year. The third thing is we are coordinating with local government to deploy this to the local government sites.
So, these are what we are currently working on. So, let me provide some idea on how this CIP is going on. And it is hard to read these icons, but these icons are most of the major news companies, including Centara and also in local area.
And also, multiple broadcasters joining on board. And also, we have a news corporation on board to help develop this technology, not only in Japan, but also in the outside of Japan.
So, this is what I want to say. And we only have 30 seconds, but if you have questions, I can answer. Thank you for that. It was really interesting. I just wondered when the user sees, like, how does the user know that when it sees this trust mark, that it, yeah, that that is trustworthy.
Like, that it sees, okay, what are you actually, who is verifying that? That's not so clear to me.
So, until we open, the user opens this pop-up on the extension icon, it will not appear. And also, do not provide any information at this moment. And we are working, some of the people, to how to make this information visible enough to the end users. But we understand that there was discussion on the EV certificate, like a green icon on the URL side. And that was not useful, so removed. We know that that's the context.
So, we are trying to find a way to present a better way to inform the end user. Thank you. If I can just quickly follow up, like, exactly how it is there. The publisher of this article has reliable information. Who is actually asserting, like, who is making that assertion? For example, in the Japan use cases, there are organizations, for example, there are newspapers and editors associations in Japan that check the policies of the newspapers.
So, we are relying on such kind of organization or certifier to provide information on the origin. That is our plan.
Sorry, just one more quick follow-up. But you're not relying on one source, but those can be multiple. Did I understand that correctly, that that was really important, that it's not one party that is providing that?
Yeah, because we do not want to provide a single control point to limit access to the entire ecosystem. So, that is the reason why we are trying to provide multiple ways to provide information. Thank you very much. I think it may have been covered by the previous question, but let's say that users get used to clicking this icon. Let's say they get used to clicking the icon, the users, and seeing this content information. What's going to stop the scammers from having a fake pop-up that says content information and users click and see it and they're happy?
Yes, that is a very important point. And we think it is important to discuss with the browser vendor to implement some way to make the information securely provided to the end user. And that is a very important point to make this technology useful.
Okay, well, thank you very much again and let's continue to our next presentation. Thank you.
