Photo/Illutration Fujitsu Ltd. and eight other entities announce a joint effort to tackle false and misleading information online at a news conference on Oct. 16. (Kanako Tanaka)

Amid the growing threat of false information fueled by the rise of generative artificial intelligence, researchers across Japan with specialized skills are working together to develop countermeasures.

Fujitsu Ltd. said on Oct. 16 that it is collaborating with universities and research institutes to create a system that verifies the authenticity of online information.

The joint project involves nine entities: Fujitsu, the National Institute of Informatics, NEC Corp., Keio University, the Institute of Science Tokyo, the University of Tokyo, the University of Aizu, the Nagoya Institute of Technology and Osaka University.

They aim to build the system by fiscal 2025, supported by 6 billion yen ($40 million) in government funding.

The system will verify information through a four-step process.

For example, when screening social media posts, the system will first convert images and videos into text to simplify its analysis and detect whether the content is a sophisticated AI-generated deepfake.

Next, it will cross-reference details such as the poster’s location, date, time and place with official announcements from the central and local governments to check for inconsistencies.

The system will then integrate these analyses to present a conclusion along with the supporting evidence.

It will also assess the reach of the post and its potential social impact.

Fujitsu plans for the system to be used by local governments and private companies, with the goal of eventually making it available to the public.

Each of the nine entities has been independently researching technology related to disinformation,

However, the rapid advancement of generative AI has led to increasingly sophisticated false information.

“Until now, Japan’s disinformation countermeasures have lagged behind those of other countries,” said Kazutoshi Sasahara, a professor at the Institute of Science Tokyo.

He said that by combining the expertise of the nine entities, they “can develop technology that leads the world.”

GOVERNMENT PUSHES LEGAL REFORMS

The government is also growing more concerned about the spread of false information online.

Following the Noto Peninsula earthquake on Jan. 1, false rescue requests spread on social media.

Fraudulent advertisements impersonating celebrities to lure people into investments have also become a serious issue.

The communication ministry convened an expert panel, which included constitutional law scholars and information technology specialists, to broadly discuss the issue of false and misleading information.

In a report of more than 130 pages compiled in September, the panel recommended comprehensive measures, including legal reforms, improving digital literacy and advancing the research, development and verification of technologies.

A new working group was set up on Oct. 10 to accelerate discussions on institutionalizing these measures.

The government plans to ask social media platform operators to take prompt action against illegal false or misleading information.

However, the ministry faces challenges dealing with false or misleading information that is not illegal. Government involvement in verifying the authenticity of such content could raise concerns about censorship.

Therefore, the ministry is focusing on creating an environment where users can assess the reliability of information on their own.

The government has allocated 440 million yen this fiscal year to support companies in their technology development.

(This article was written by Kanako Tanaka and Naoko Murai.)

Related News