WAN-IFRA Media Congress / Safer Digital Experience: Presentation Extracts By Yomiuri Managing Editor Riichiro Maeki

The Yomiuri Shimbun
Yomiuri Shimbun Managing Editor Riichiro Maeki delivers a speech at the World News Media Congress in Taipei on Friday.

The following is excerpted from a presentation given by Riichiro Maeki, managing editor of The Yomiuri Shimbun, on June 30 in Taipei during the 74th World News Media Congress organized by the World Association of News Publishers (WAN-IFRA).

People around the world are trying to find ways to solve concerns regarding generative AI. The most critical issue I see concerns the sources of information.

All of you here today working in newspapers and other media are truth seekers who place importance on sources. It is imperative that we check the sources to determine the accuracy of facts.

Generative AI consumes vast amounts of information on the internet to generate answers, but it is not entirely clear where it gets its sources. Concerns about whether generative AI might be misused to spread false information must be an issue that cannot be ignored.

With the advent of the digital age, newspapers and many other news organizations have had to deal with new challenges. We must now clarify where information comes from and how reliable it is.

There is no room for any falsehood in what we carry in The Yomiuri Shimbun. Only what our reporters actually see and hear are included in our firsthand information. Generative AI, which creates output based on existing secondhand information, is alien to journalism. Nevertheless, under these circumstances, even The Yomiuri Shimbun’s coverage must clarify that it is authentic.

Today, I would like to introduce a new initiative in Japan. My country’s newspapers, TV stations and other media, advertising firms, a telecommunications giant and a platform enterprise have come together to improve the situation. What we have come up with is a framework based on a technology called Originator Profile, or OP.

The goal of the OP is to reveal the source or originator of information. This is done just like embedding watermarks in banknotes. The OP embeds an electronic identifier in each piece of information, such as an article or advertisement, to guarantee its authenticity.

Specifically, data about the sender of information is embedded in the identifier, allowing users to view information related to credibility, which is confirmed by third-party organizations. The credibility-related information means, for example, stating the editorial policy and reporting responsibility for media companies and the corporate stance for advertisers. We would like to embed this information in each article, so that it can function when the article is shared over social media.

Disinformation can spread online under the guise of news articles because it is possible to disseminate them without revealing sources. In addition to the human eye, we also need technology and mechanisms that can determine the source of information mechanically and instantly.

This is a matter of life and death for our business model, where we painstakingly produce articles every day to generate revenue by spending a lot of money, time and effort for the dissemination and distribution of correct information. With the use of an app for the OP, people browsing the internet can check the sources of information when they come across something suspicious. Once such a mechanism is established, the internet will be a safer space.

No matter how excellent the technology, however, if many people do not use it, the internet will not be a safer space.

Therefore, we have decided to create an organization called the Originator Profile Collaborative Innovation Partnership (OPCIP) in order to include as many of the various stakeholders in the media and advertising in Japan as possible for the development of the technology.

The OPCIP was established in December 2022. It has 27 member companies and organizations as of today. All of Japan’s national newspapers are members of the partnership. TV stations, major advertising companies, a telecommunication giant and a platform enterprise are all part of the OPCIP. These companies have teamed up to test the OP technology on their own websites, with the goal of putting it into practical use.

We will welcome the involvement of advertising companies and advertisers in the future because the OP will be useful also for advertising. We believe that the technology can help provide a solution to issues such as ad fraud, when the number of clicks on ads are inflated by bots, and advertisers get charged more. It will also help protect brands from the risk of being damaged by having their advertisements posted on illegal websites.

We will start testing the technology and hope to try a system in which digital advertisements embedded with the electronic identifiers are published along with articles with similarly embedded identifiers, recognizing each other, in an open internet environment by the end of the year. We aim for the widespread use of the OP in Japan by the end of 2025.

We do not intend to limit our efforts to Japan alone. We aim to bring it to the World Wide Web Consortium (W3C), which determines international agreements on the internet, and to have it adopted as an international standard. Our goal is to contribute to a healthy internet.

Getting back to the concerns with generative AI that I mentioned at the beginning, at present, the technological progress of generative AI is advancing without the development of rules. If the use of generative AI continues at this rate, the OP technology to clarify the source of information and ensure the authenticity of articles and advertisements might be undermined, and the harmful effects of fake news and false information might increase.

Countries around the world are working on the development of rules concerning generative AI. In Europe, there seem to be moves to revise laws to take copyright into consideration, such as requiring the disclosure of materials used to instruct generative AI. However, there are concerns that Japan is rather negative about making such rules in the belief that the nation must not lag behind in the competition for developing AI technology.

The Yomiuri Shimbun will continue to point out various risks associated with generative AI, such as the spread of false information and copyright infringement, and call for generative AI to be used in a well-balanced way.