Claude, War, and the State of the Republic
The clash between the artificial intelligence company Anthropic and the Department of War is about far more than technology. At its core, it reveals a dangerous and accelerating expansion of executive power in the United States — one that threatens the constitutional structure that restrained government authority for over two centuries. Based on a discussion between Russ Roberts and Dean Ball on EconTalk, the dispute illustrates how modern administrations are increasingly willing to use the power of the federal government not merely to govern, but to pressure, punish, and economically cripple private actors who refuse to comply with political demands.
The controversy began when Anthropic refused to remove contractual restrictions that prevented its AI model Claude from being used for domestic mass surveillance and autonomous lethal weapons systems. In response, the Department of War reportedly threatened to classify the company as a “supply chain risk,” a designation typically used against hostile foreign entities. Such a move would have severely damaged Anthropic’s ability to conduct business with military contractors and could have crippled the company financially.
What makes this incident so alarming is not simply the disagreement itself, but the mechanism of coercion being used. The federal government was not attempting to pass legislation through Congress. It was not seeking a constitutional amendment or engaging in open democratic debate. Instead, executive agencies leveraged regulatory power and economic pressure to compel compliance from a private company. That distinction is critical. It represents the transformation of executive power from administration of the law into direct political force.
This trend did not begin with one administration or one political party. The growth of executive authority has been unfolding for decades. During the presidency of Franklin D. Roosevelt, the federal government expanded dramatically through the New Deal, creating administrative agencies with broad regulatory powers that were never envisioned by the Founding Fathers. While many Americans supported those programs during the Great Depression, the long-term consequence was the normalization of a permanent bureaucratic state capable of issuing rules with the force of law without direct congressional approval.
The expansion accelerated during the Cold War. National security concerns allowed presidents to justify extraordinary executive actions in the name of protecting the nation. The rise of intelligence agencies, classified operations, and emergency powers gradually shifted authority away from Congress and toward the executive branch. By the time of the Vietnam War and the Watergate scandal, Americans were already beginning to recognize how dangerous unchecked executive authority could become. President Richard Nixon famously suggested that when the president acts in the interest of national security, legality becomes secondary. That mentality permanently altered public trust in constitutional limits.
Following the terrorist attacks of September 11, 2001, executive power expanded once again under the justification of security and emergency response. The passage of the USA PATRIOT Act dramatically increased surveillance capabilities and government access to private information. Intelligence agencies gained powers that previous generations of Americans would have considered unthinkable. Yet many citizens accepted these changes because they believed the threats justified them. History repeatedly demonstrates that governments rarely surrender powers once they acquire them.
The Anthropic conflict reflects the next stage of this progression. The emergence of artificial intelligence gives governments unprecedented potential to analyze data, monitor populations, and automate decision-making at scales previously impossible. As Dean Ball explained, modern AI systems could allow governments to process commercially available data on millions of citizens simultaneously, effectively creating surveillance capabilities far beyond anything existing law anticipated.
This is precisely why the executive branch’s response is so concerning. Instead of acknowledging the need for new constitutional and legal safeguards, officials attempted to force a private company into unconditional cooperation through economic intimidation. The danger lies not only in what one administration may do, but in the precedent being established. Once the federal government demonstrates that it can economically punish companies for refusing political demands, the barrier separating public authority from private enterprise begins to disappear.
Historically, one of America’s greatest advantages over authoritarian systems was the independence of private institutions from the state. In countries such as the former Soviet Union or modern China, major corporations operate with the understanding that they ultimately serve government interests. Roberts and Ball warned that America risks moving toward a similar model, where companies survive not by innovation or merit, but by political obedience.
The most troubling aspect is how normalized this process has become. Americans increasingly evaluate government actions based on whether they achieve desired outcomes rather than whether they remain constitutionally restrained. The Constitution was designed specifically to slow government power and make coercive action difficult. The Founding Fathers feared concentrated executive authority because they understood how quickly republics can erode into systems governed by discretion rather than law.
Benjamin Franklin allegedly remarked after the Constitutional Convention that the United States had created “a republic, if you can keep it.” That warning feels increasingly relevant today. The danger facing modern America is not likely to arrive through a sudden dictatorship or dramatic overthrow of democracy. It emerges gradually through the erosion of institutional limits, the expansion of executive discretion, and the public’s growing acceptance of government coercion when used against political or ideological opponents.
The Anthropic dispute may appear at first glance to be a technical disagreement about artificial intelligence contracts. In reality, it is part of a much larger historical pattern: the steady transfer of power away from constitutional institutions and toward executive authority. If left unchecked, that process threatens not only private enterprise, but the very structure of the American republic itself.