The next battlefield in the AI race

Shim Jae-hoon

The author is a foreign attorney at Hyemyung Law Firm and adjunct professor at KAIST's Moon Soul Graduate School of Future Strategy.

Advanced technology has long been celebrated for breaking down borders. Today, it is drawing new ones.

The recent decision by the U.S. Department of Commerce to restrict access to certain AI services sent a clear warning to the global technology industry. Officially described as service restrictions, the measures effectively established technological borders. By excluding not only users outside the United States but also foreign-national employees of affected companies, Washington demonstrated that AI is no longer a freely shared technology. Code and algorithms, once regarded as commercial assets, have become strategic resources that directly influence national security.

The move is the broadest U.S. technology control since U.S. President Franklin D. Roosevelt signed the Export Control Act in 1940. More importantly, it marks a fundamental shift in policy. Previous controls focused on tangible assets such as weapons, raw materials and semiconductors. The latest restrictions extend national security oversight to general-purpose AI models used in everyday applications. AI, long viewed as civilian technology, has moved to the center of geopolitical competition.

Attention is now turning to where the Commerce Department's next restrictions may fall. The most likely target is physical AI, which combines artificial intelligence with robots, drones and autonomous vehicles, allowing machines to perceive and interact with the real world. This is where digital intelligence converges with manufacturing capability, making it the next major battlefield in the AI race.

If competitiveness in the new economy can be divided into brains and muscles, Korea has built considerable strength in the former. Its semiconductor companies remain globally competitive in memory chips and advanced semiconductors that power AI systems. However, the country still trails in physical AI, the industrial muscle that turns intelligence into action.

Success in physical AI depends on core technologies such as precision sensors, high-performance actuators and reduction gears that enable humanoid robots to move accurately. This field offers Korea an opportunity to transform its manufacturing expertise, semiconductor leadership and battery technologies into a new growth engine.

The United States is expanding its AI strategy beyond software to the hardware that connects AI with the physical world. Before tighter restrictions emerge, Korean companies should use their manufacturing strengths to build a competitive physical AI ecosystem. The contest for AI leadership is shifting from algorithms to hardware, and Korea has little time to decide on which side of that technological frontier it will stand.