8:00 JST, September 5, 2026

Debate over AI sovereignty is intensifying in Japan. Previously, the focus was on developing and possessing domestic AI models, but it has recently shifted toward ensuring access to AI models in cooperation with allies and like-minded countries.

The export controls imposed by the U.S. government on the American company Anthropic in June were a turning point in this discussion. Specifically, the U.S. government banned the provision of Anthropic’s cutting-edge AI models — Claude Mythos 5 and Claude Fable 5 — to foreign nationals, citing their advanced ability to identify system vulnerabilities and the risk that they could be misused in cyberattacks.

Although the restrictions were subsequently lifted, the move sent shockwaves through Japan and Europe, where AI development has relied heavily on access to cutting-edge U.S. models. It became starkly apparent that AI is a strategic technology subject to state control, making AI sovereignty a critical national priority for countries around the world.

AI sovereignty extends beyond AI models such as large language models (LLMs). It also encompasses the power needed to run AI systems and data centers, GPUs for computing, infrastructure such as cloud services and networks, data for machine learning, and the applications and interfaces used by end users. The concept is analogous to securing supply chains from an economic security perspective.

The Japanese government’s vision for AI sovereignty is articulated through the concept of “open AI sovereignty,” as outlined in the AI Basic Plan revised on July 14. The plan states that........

© The Japan News