Large-scale data breaches have occurred one after another at companies and local governments in Japan, with cases of attacks using AI confirmed. This article organizes the number of records exposed, specific examples of damage, and the key points of companies' countermeasures.
The number of publicly disclosed incidents in Japan has been increasing on a daily basis since the end of September, and the government is accelerating its response, including by introducing emergency measures. According to figures compiled by Trend Micro, the number of incidents disclosed this year has already exceeded the total for the previous year. Damage including data breaches and service outages has spread widely, prompting companies to reassess their security measures. While the basic pattern of attacks remains the exploitation of vulnerabilities and targeting of inadequate management, their scale and speed have increased substantially.
Specific examples of damage include unauthorized access to approximately 6.6 million accounts at major car-sharing service Times Car and 10,788,963 member records at Yakiniku King. IDCフロンティア's cloud infrastructure was also brought down by ransomware, affecting 495 companies and local governments. In addition, GMO Internet Group set a bounty of up to $1 million for information leading to the arrest of the perpetrators.
JPCERT/CC's alert has revealed that attackers were using AI as an active intrusion tool. It is currently unclear which breaches involved AI, but observers have pointed to the possibility that self-directed workflows known as AI agents are reducing the need for human involvement in mechanical tasks such as reconnaissance and exploration, thereby increasing the scale and tempo of attacks.