Photo/Illutration The Defense Ministry’s recent artificial intelligence policy was accompanied by a strategy to create a specialized workforce well-versed in AI. (Asahi Shimbun file photo)

The Defense Ministry has hashed out a basic policy to promote the use of artificial intelligence and introduce generative AI to bolster national defense.

The policy outline, which is the first of its kind, states a clear need for “human involvement” in making decisions on actual operations of the Self-Defense Forces. It also reiterates the government stance that Japan has no intention to develop “fully autonomous lethal weapons.”

Released on July 2, the document specifies seven priority fields, describing how an AI system could assist the decision-making process during complicated operations.

These seven areas are target detection and identification; information gathering and analysis; command and control; logistics support; “unmanned assets” (equipment), such as drones; cybersecurity; and clerical work.

In some cases this could mean using AI systems to identify targets based on radar tracking and satellite images. Another hope is to analyze vast volumes of online information and assist commanders in making prompt decisions.

Some tout it as game-changing technology that will help speed up combat operations and reduce human errors while saving on manpower.

As national defense encompasses land, sea, air, space, cyberspace and electromagnetics--among other spaces--the document cites “cross-domain operations” as a concrete example of exactly how an AI system would be beneficial.

The said system could work as a tool to aid in drafting an action plan for performing operations under integrated command and control by linking various equipment into a network.

The document also included a plan to work toward introducing generative AI, which is undergoing rapid technological development.

The ministry also recognizes that AI overall is not a magical solution and has limitations.

The document also acknowledges that, as these systems learn based on existing data, they would be less useful in unprecedented situations as they “lack the capacity to locate what the problems are.”

It envisions assigning experts to identify which questions need troubleshooting and wielding these systems in fields where there is ample quality data to train the AI.

The ministry has also worked out a comprehensive strategy to develop a specialized workforce for AI systems.

Introduction of a career track where SDF recruits focus on cyber-related tasks upon enrollment is part of this strategy. From fiscal 2025, the Ground SDF plans to introduce an employment test exclusively for personnel with cyber capabilities.

Related News