Fujitsu Leads Consortium to Combat Disinformation with Cutting-Edge Platform
Kawasaki, Japan – October 16, 2024 – In an era of rampant misinformation and sophisticated manipulation tactics, Fujitsu, a global leader in information and communication technology, has embarked on a mission to develop a groundbreaking disinformation countermeasure platform. This ambitious project, backed by Japan’s New Energy and Industrial Technology Development Organization (NEDO) under the Key and Advanced Technology R&D through Cross Community Collaboration Program (K Program), aims to create the world’s first comprehensive solution to tackle disinformation from detection to analysis and evaluation. The platform is slated for completion by the end of fiscal year 2025.
Fujitsu has assembled a formidable consortium of academic and private sector powerhouses, including the National Institute of Informatics (NII), NEC Corporation, Keio Research Institute at SFC, Institute of Science Tokyo, The University of Tokyo, University of Aizu, Nagoya Institute of Technology, and Osaka University. This collaborative effort brings together leading experts in diverse fields, from artificial intelligence and media analysis to cognitive science and computational social science. Vivek Mahajan, Corporate Vice President, CTO, CPO of Fujitsu Limited, emphasized the significance of this initiative, highlighting the consortium’s proven track record in combating disinformation and the collaborative spirit driving the project.
The project will leverage a multi-pronged technological approach. First, NII and NEC will spearhead the development of sophisticated information analysis and disinformation detection technologies. NII will focus on detecting manipulated media like deepfakes, assigning confidence scores to detected manipulations, while NEC will develop media understanding technology to extract textual information from various media formats and analyze its correlation with social media posts to gather supporting evidence. This combined approach will provide a robust initial layer of defense against disinformation.
Second, Keio Research Institute at SFC, Fujitsu, and Osaka University will collaborate on evidence and endorsement management. Keio Research Institute at SFC and Fujitsu will develop technology to integrate and structure the collected evidence into an evidence/endorsement graph, a novel approach to verifying authenticity and assessing impact. Osaka University will contribute its expertise in IoT data analysis to gather sensor data, even inferring evidence from neighboring areas when data is scarce, simulating human information gathering processes.
Third, Fujitsu and Nagoya Institute of Technology will address the challenge of comprehensive authenticity determination. Fujitsu will develop technology to analyze the consistency of evidence using the evidence/endorsement graphs and create a specialized Japanese Large Language Model (LLM) for disinformation countermeasures. This specialized LLM, leveraging Fujitsu’s supercomputer (Fugaku) and LLM (Takane) expertise, will enhance comprehension of news and social media data, improve logical reasoning, and accelerate high-accuracy authenticity verification while mitigating hallucinations. Nagoya Institute of Technology will focus on developing a user-friendly interface based on cognitive science principles, assisting users in accurately judging information veracity and discouraging the spread of false information.
Fourth, the Institute of Science Tokyo, The University of Tokyo, and University of Aizu will tackle the evaluation of disinformation impact. These institutions, with their deep expertise in computational social science, will develop an advanced AI model for disinformation assessment, extending the capabilities of LLMs. By analyzing various characteristics of disinformation, such as similarity to past campaigns and proliferation speed, focusing on the source, content, and social context of messages from social media data, they will evaluate metrics like proliferation scale and social impact, providing a quantitative assessment of disinformation’s effect on society.
The project is structured in phases. Fiscal year 2024 will focus on analyzing use cases for private and public sector organizations, defining functional requirements, and commencing research and development of individual technologies. The integration of these technologies into a fully functional platform is scheduled for completion by the end of fiscal year 2025. The consortium aims to contribute to economic stability by mitigating the risks posed by disinformation, fostering new industry development, enhancing international competitiveness, and strengthening Japan’s global standing.
The consortium partners expressed enthusiasm and commitment to this groundbreaking initiative. Prof. Junichi Yamagishi of NII emphasized the importance of an integrated technological approach in combating disinformation. Motoo Nishihara, NEC’s CTO, highlighted the necessity of AI-powered countermeasures to address AI-generated disinformation. Keio University professors highlighted the importance of tracking sender identity and information relationships using cutting-edge digital identity technology. Prof. Kazutoshi Sasahara of Institute of Science Tokyo stressed the urgency of developing tools to assess disinformation’s influence. Prof. Masashi Toyoda of The University of Tokyo emphasized the collaborative nature of the project and its potential to build a robust infrastructure for assessing disinformation’s impact. Yasuhiro Hashimoto of Aizu University highlighted the multidisciplinary approach needed to combat this complex threat. Prof. Yuko Tanaka of Nagoya Institute of Technology focused on the importance of a user-friendly interface designed with cognitive science principles in mind. Finally, Shin’ichi Arakawa of Osaka University emphasized the practical focus of their research and development efforts. With such a diverse and dedicated team, Fujitsu’s disinformation countermeasure platform promises to be a game-changer in the fight against misinformation.