A Google expert describes the front line of AI-enabled cyberattacks and the broader picture of incidents targeting Japanese companies. Attackers are said to use a range of AI models, including GPT, Claude, and Gemini, with AI potentially serving as a means of gaining entry and stealing resources.
The cutting-edge program “1on1 Tech” covered the front line of AI-enabled cyberattacks through an interview with John Hultquist, chief analyst of Google's Threat Intelligence group. The program focused on how attackers are using AI and how companies and individuals should prepare.
According to the interview, attackers are beginning to combine multiple AI models and use AI as agents that carry out attacks, moving beyond translation and technical advice. Cases of attackers relying on Gemini were also cited, and attack methods were reported to be becoming more diverse.
Amid a succession of cyberattacks and information leaks affecting Japanese companies, it was pointed out that everyday devices such as home routers can also serve as stepping stones for attacks, and that viewers should recognize the need to strengthen preparations for both their companies and themselves. The program explained that the approach to defense is also changing as AI advances rapidly.
The program's guests included Nakagawa Masahiro (中川雅博), business editor at TBS CROSS DIG with Bloomberg, and Hultquist. The recording date is explicitly given as 20 October 2026.