Opinion | The AI War Has Already Begun—and Trump Just Picked a Side

“The radical left at Anthropic has made a grave mistake, attempting to force the Department of War to disregard our Constitution and comply with their terms of service. Their selfishness puts the American people at risk, endangers our troops, and threatens our national security… Therefore, I order all agencies of the U.S. federal government to immediately cease using Anthropic's technology. We don't need it, don't want it, and will never do business with this company again!”

—Donald Trump, President of the United States.

“I firmly believe that AI is indispensable for defending America and other democratic nations, and for defeating authoritarian adversaries.”

—Dario Amodei, Co-founder and CEO of Anthropic

A Sudden Rupture in Silicon Valley

For much of his presidency, Donald Trump has been broadly supportive of the generative AI industry, easing regulatory pressure while encouraging government adoption of emerging technologies. That relationship changed abruptly on February 27, when he ordered U.S. federal agencies to stop using technology from Anthropic, one of Silicon Valley's most prominent AI startups.

The order effectively blacklisted the company from federal procurement and threatened a defense contract reportedly worth around $200 million. Within hours, the decision exposed a deeper conflict—one that goes beyond a single company and reflects a growing divide over how artificial intelligence should be used in modern warfare.

At the center of that debate is Anthropic's large language model, Claude.

From Chatbot to Battlefield System

Claude emerged in 2023 as one of the most advanced generative AI models, named after information theory pioneer Claude Shannon. Its product versions—Haiku, Sonnet, and Opus—suggest a creative lineage, but the system has increasingly moved far beyond consumer chatbot applications.

In recent years, AI models have been integrated into defense data platforms designed to process enormous volumes of intelligence. Systems connected to the Pentagon's Project Maven architecture, for example, help analyze surveillance imagery, detect potential targets, and generate recommendations for military planners.

Such platforms do not replace human commanders, but they dramatically accelerate decision-making. Instead of analysts manually compiling intelligence for hours or days, machine-assisted systems can process vast datasets in minutes.

Paul Scharre, a military technology specialist, has described the shift succinctly: modern AI allows militaries to generate operational insights “at machine speed rather than human speed.”

That change may prove as consequential as earlier revolutions in radar, precision weapons, or satellite reconnaissance.

The Limits of Machine Intelligence

Yet AI systems remain far from infallible. Even their most enthusiastic advocates acknowledge that automated analysis can produce errors, misinterpret data, or amplify faulty assumptions embedded in training information.

In a battlefield environment, those risks carry enormous consequences. A flawed assessment can lead to mistaken targeting decisions or escalate conflicts based on incomplete information.

For this reason, debates about military AI increasingly focus not only on technological capability but also on governance—specifically, how much autonomy machines should have in life-and-death decisions.

Anthropic has attempted to draw a clear line. （Related： Chiang Kai-shek's Long Shadow: What Rubio's Cold War Obsession Reveals About Taiwan ｜ Latest ）

The Line Anthropic Refused to Cross

Unlike many technology startups, Anthropic is structured as a public-benefit corporation. Its stated mission emphasizes building AI systems that remain understandable, controllable, and aligned with human oversight.

In negotiations over a new Pentagon contract, the company reportedly asked the U.S. military to accept two limitations on the use of its technology: no deployment for mass domestic surveillance and no use in fully autonomous weapons systems operating without meaningful human control.

Anthropic has argued that AI can strengthen democratic defense while still respecting civil liberties and ethical constraints. But such conditions collided with the Pentagon's traditional procurement philosophy.

From the military's perspective, private companies deliver technology—they do not dictate operational policy. Decisions about how weapons systems are used ultimately belong to elected governments and military commanders.

Politics Enters the Battlefield

The dispute might have remained a technical disagreement over procurement rules. Instead, it rapidly became politicized.

Trump framed the conflict as part of a broader ideological struggle between his administration and what he described as politically progressive technology companies. His order to halt federal use of Anthropic systems turned a contract dispute into a public confrontation.

The timing was significant. On the same day the ban was announced, OpenAI revealed that it had reached a new agreement with the Pentagon to provide AI systems for government use under existing legal frameworks.

The contrast was stark. Where Anthropic had sought explicit limitations, OpenAI signaled a willingness to work within government guidelines while leaving operational decisions to defense authorities. （Related： Chiang Kai-shek's Long Shadow: What Rubio's Cold War Obsession Reveals About Taiwan ｜ Latest ）

The Emerging AI Arms Race

Behind the political drama lies a larger transformation. Artificial intelligence is becoming an integral component of military planning, logistics, intelligence analysis, and targeting support.

Modern warfare increasingly depends on data—satellite imagery, sensor feeds, communications intercepts, and battlefield reports. AI systems capable of synthesizing this information quickly offer a powerful strategic advantage.

As more governments adopt similar technologies, the role of AI in conflict is likely to expand rapidly. The result could be a global competition to develop faster, more capable military algorithms.

In that sense, the dispute between Trump and Anthropic may represent an early skirmish in a much broader struggle over how AI will shape the future of war.

No Easy Answers

The controversy also highlights a fundamental dilemma. Democracies rely on technological innovation to maintain security, yet those same innovations raise ethical questions that cannot be easily resolved.

Should private companies be allowed to impose moral limits on how governments use their technologies? Or should decisions about military AI rest entirely with elected leaders and national defense institutions?

Neither approach offers a perfect solution.

What is increasingly clear, however, is that artificial intelligence is no longer a distant research frontier. It has already become part of the strategic infrastructure of modern states.

And as the confrontation between Washington and Silicon Valley shows, the political battles over AI may prove almost as consequential as the technology itself.

*The author is a veteran media professional and professional translator.

（Related： Chiang Kai-shek's Long Shadow: What Rubio's Cold War Obsession Reveals About Taiwan ｜ Latest ）