TOKYO - Japan is considering utilizing cutting-edge artificial intelligence systems under its preemptive cyber defense policy in the face of the risk of attacks by advanced models such as U.S. startup Anthropic's Claude Mythos, a senior government official said.
Amid growing concerns over cyberattacks by AI models with high capabilities to detect security vulnerabilities, it is "inevitable" that Japan will use sophisticated AI technology, said National Cyber Director Yoichi Iida.
"It would be no surprise if the time between vulnerability discovery and exploitation were shortened to a fraction of a hundredth or a thousandth of what it used to be," Iida said in a recent interview with Kyodo News.
The Claude Mythos model is seen as highly capable of finding security vulnerabilities, and Anthropic initially limited its access to IT companies such as Google LLC and certain financial institutions after its April unveiling, due to concerns it could be misused for cyberattacks.
Iida's remarks came after Prime Minister Sanae Takaichi's Cabinet decided in March to enable the government from Oct. 1 to take "active cyber defense" measures to eliminate in advance the possibility of serious cyberattacks that could disrupt critical infrastructure, economic activity and people's livelihoods.
The government needs to consider "how to advance automation" in its cyber defense operations while ensuring human involvement in the final decision-making process, Iida said. He stopped short of clarifying possible concrete steps.
Iida also said Japan should not rely on any one particular AI model, citing an example of the U.S. government's export control directive to suspend access to Mythos-class models in June.
The suspension affected security vulnerability scanning operations for key government systems, Iida said.
If cutting-edge AI models were exploited by nation-sponsored cyberattacks, such offensives could become "far more sophisticated," and their intensity is expected to increase, Iida added.
