Japanese Consortium Develops AI Fact-Checking Platform to Combat Online Misinformation

A Japanese consortium, with Fujitsu Ltd. as a key participant, is currently working on a fact-checking platform powered by artificial intelligence. The primary goal of this system is to combat the spread of online misinformation, particularly during critical events like natural disasters and elections.

The consortium comprises nine organizations, including the National Institute of Informatics, NEC Corp., and various academic institutions. The target completion date for the system is the end of fiscal year 2025. The AI will be designed to analyze information found online, gather supporting data, and ultimately assess the authenticity of the claims.

According to Dai Yamamoto, a senior project director at Fujitsu, the system is intended to streamline the process of verifying information. He highlighted that the AI could facilitate quick judgments, which would be a significant improvement over the time-consuming manual fact-checking process.

During a recent test of the system, Yamamoto input the statement, "A group of foreign thieves went to quake-hit areas immediately after the Noto earthquake," to test its verification capabilities. The system responded within seconds, labeling the statement as "False."

The system's assessment was supported by a newspaper article that contradicted the initial claim. Furthermore, the system evaluated the reliability of the news organization that published the article, concluding it was "high."

The development of this platform is being funded by the New Energy and Industrial Technology Development Organization, a government agency, with an investment of 6 billion yen ($40 million). The system utilizes a large language model specifically designed to combat false information.

Developers also claim the platform will be able to detect "deepfakes," which are AI-generated images and videos that can feature real people. The system is designed to identify subtle characteristics often missed by human observers.

This initiative is a response to growing concerns about the proliferation of online disinformation and misinformation, including fake videos related to disasters, and their potential negative consequences during times of crisis.