Japan seeks AI system for predicting cyberattacks

TOKYO — The Japanese government will undertake a research and development project to use artificial intelligence to automatically detect signs of cyberattacks and evaluate their potential impact, The Yomiuri Shimbun has learned.

The hope is that sharing early-warning information among relevant institutions will lead to rapid responses. Work could start within the month, with plans to conduct verification experiments in fiscal 2022 and have working technology as early as possible.

R&D will focus on developing technology to analyze online data to detect the early stages of malware attacks. The idea is to use the deep-learning capabilities of AI on massive amounts of observational data to automatically detect attacks.

The government also plans to develop technology that can quickly analyze the characteristics and impacts of malware attacks that are detected to obtain early-warning information on cyber-attacks. This warning information would be sent to government agencies, private companies and other relevant entities.

The government hopes that once the technology is established, it will prevent or minimize damage to targeted institutions.

According to the Internal Affairs and Communications Ministry, although technologies for detecting attacks from past patterns exist, an AI system that predicts future attacks using deep learning has yet to be established. If successful, such a system would be cutting-edge.

R&D is to be carried out in collaboration with private companies and research institutes, led by the National Institute of Information and Communications Technology, which is under the ministry’s jurisdiction. Work this fiscal year will focus on tasks such as accumulating data on attacks, with full-scale technological development to begin next fiscal year.

The policy on using AI to deal with cyberattacks was laid out in “organizational guidelines on cybersecurity research and technological development,” which was drafted in May by a specialist inquiry board of the government’s cybersecurity strategic headquarters, which is headed by Chief Cabinet Secretary Yoshihide Suga.

Cyber-attacks have been on the rise recently. As new types of malware are constantly appearing, the threat is only expected to grow. Cyber-attacks that put AI to malicious use are also expected.

With the rapid spread of products and services connected to the internet of things, the government wants to strengthen cooperation with industry and academia to improve cyberdefenses.