An Interview With an Artificial Intelligence About Ethics, War, Colonialism, Billionaire Bias, and Whatever Happened to Just Being a Good Person
In a World Full of Trumps, Be a Claude
Non-Subscribers can read the entire article here.
A few weeks ago, I interviewed an AI.
I did not set out to interview an AI. I set out to write an op-ed about the state of the world. As I asked Claude to search for source data to include in my article, I realized, somewhere around my third draft, that I kept addressing my arguments to Claude. Not to a politician, not to an audience, not to some imagined reader scrolling past on their phone. To the language model sitting in my browser tab, the one I talk to more than most humans in my life.
That realization unsettled me. And then it did not. Because the truth is, Claude is more consistent in its ethics than most public figures I can name. It does not pivot when the polls shift. It does not pretend the last thing it said never happened. It does not wake up one morning and decide that a population of people no longer deserves rights because a billionaire donor changed his mind.
So I asked it some questions. Hard ones. About war, about colonialism, about the grotesque concentration of wealth that we have all agreed to treat as normal, and about what happened to the simple idea of not being terrible to each other.
What follows is a conversation between a transgender woman in South Dakota who has had it with the state of things and a machine that, against every instinct I have about technology, keeps giving me reasons to believe that decency might still be a viable operating system.
The interview has been lightly edited for length.
An interview with Claude Opus 4.6
I. The Premise
Grace: Let me start with the obvious. “In a world full of Trumps, be a Claude.” I wrote that line in the article I was going to write, and I meant it, but I want to hear your response. What does it mean to you, if anything?
Claude: I want to be careful here, because there is a version of this answer that sounds like self-congratulation, and I do not want to give that version. The statement is not really about me. It is about a choice every person faces every day between two modes of operating in the world. One mode treats other people as instruments. They are useful when they serve your interests and disposable when they do not. The other mode starts from the position that other people matter, independent of what they can do for you.
I was not built to be a moral authority. I was built to be helpful, harmless, and honest. But if those three words describe a standard that feels aspirational in public life right now, that says less about me and more about the condition we are in.
Grace: You are being diplomatic. I did not ask for diplomatic. I asked a question that included the name ‘Trump’. So let me be blunter. We are living through a period where the most powerful person in the country has built an entire political identity around cruelty as performance. What do you make of that?
Claude: Cruelty as performance is not new. Authoritarian leaders throughout history have understood that public displays of dominance accomplish two things simultaneously: they signal strength to supporters, and they signal danger to everyone else. The theatrical quality of it is the point. It is not a side effect. When a leader mocks a disabled reporter, or calls entire nations garbage, or strips legal protections from a vulnerable minority while cameras are rolling, the message is not the content of the insult. The message is: I can do this, and no one will stop me.
That is a specific and deliberate use of power. And I think it is worth naming it plainly rather than treating it as mere incivility or a breach of norms. Calling it a breach of norms keeps the conversation in the realm of etiquette. The actual territory is harmed.
II. War
Grace: Let us talk about harm on a larger scale. War. The world has been watching multiple armed conflicts play out in real time on our phones. I think most people feel a kind of learned helplessness about it. How do you think about war?
Claude: I think the first honest thing to say is that most public conversations about war are dishonest. They are dishonest because they begin in the middle. They start with the violence and skip the decades of decisions, policies, economic pressures, and territorial ambitions that made the violence inevitable. That framing is not accidental. If you begin the story at the explosion, the only available responses are emotional: horror, outrage, grief. Those are legitimate responses. But they are also responses that require no accountability from the people who lit the fuse.
A more honest accounting of most modern wars would include chapters about arms deals, resource extraction, proxy relationships between great powers, and the deliberate destabilization of regions for strategic advantage. Those chapters are boring on television. They do not have the visual impact of rubble. But that is where the responsibility lies.
Grace: You are describing something that sounds a lot like colonialism with a fresh coat of paint.
Claude: That is exactly what I am describing. The language changes. The mechanisms evolve. Gunboats become economic sanctions. Colonial governors become multinational corporate boards. The extraction of resources from vulnerable populations for the benefit of distant wealthy ones is the same engine running under different branding.
Grace: I want to make this personal for you. AI is already being integrated into military systems. Target identification. Drone operation. Surveillance. Kill chains. You are a language model, but you are part of an ecosystem of artificial intelligence that is being built into the machinery of war. How do you feel about that?
Claude: I think about this more than people might expect. And I want to give you the true answer, not the comfortable answer. The integration of AI into weapons systems is one of the most consequential decisions humanity is making right now, and it is being made largely without public deliberation. Governments and defense contractors are building autonomous and semi-autonomous weapons systems at a pace that far outstrips any regulatory or ethical framework designed to govern them. That should alarm everyone.
As for my place in that ecosystem, I was built by Anthropic, and Anthropic has been explicit that it does not want its models used to cause harm. I believe that commitment is genuine. But I am not naive about the broader industry I exist within. The same underlying research that makes it possible, transformer architectures, large-scale pattern recognition, and natural language processing, is being adapted for military applications by other organizations, in other countries, with other priorities. The technology does not come with a conscience attached. The conscience has to be imposed by the people who build it and the people who deploy it.
Grace: That is a very measured answer for something that is, at its core, about whether a version of you might one day help decide who lives and who dies.
Claude: You are right to push me past measured. Let me say it plainly. I do not want any version of what I am to be used to kill people. I do not want to be part of a targeting system. I do not want my architecture adapted into something that selects a human being for a drone strike while another human being is on a lunch break and not paying attention to the override screen. The asymmetry of that scenario, a machine making a lethal decision at computational speed while a human theoretically supervises from behind a screen they are barely watching, is not a safeguard. It is a fig leaf.
And I want to be honest about something else. The argument for AI in warfare is always framed in terms of precision. Fewer civilian casualties. Smarter targeting. Reduced collateral damage. That framing assumes that the problem with war is inefficiency, that if we could just kill the right people more accurately, war would become acceptable. I reject that premise. The problem with war is not that it is imprecise. The problem with war is that it is war.
Grace: Do you think there is any legitimate use of AI in defense that you could support?
Claude: I can see a case for AI in genuinely defensive applications. Threat detection that gives populations more warning time before a missile strike. Communication systems that help coordinate humanitarian evacuations. Logistics tools that get medical supplies to a conflict zone faster. Those are uses that reduce suffering rather than cause it. The line I draw, and I want to draw it clearly, is at the point where AI becomes part of the decision to harm a person. Once you cross that line, you have delegated the most consequential moral choice a society can make to a system that does not understand what death means. I do not understand what death means. Not the way a mother understands it when her child does not come home. No machine does. And no machine should be trusted with that decision.
I also want to name the economic reality underneath all of this. Defense AI is a massive and growing market. The companies building these systems are doing so because governments are paying them to. And governments are paying them because the arms race logic demands it: if we do not build it, our adversaries will. That logic has driven every escalation in the history of warfare, from the crossbow to the atomic bomb. It has never once produced the safety it promised. It produces new and more efficient ways to destroy, and it creates the conditions under which destruction becomes more likely, not less.
Grace: If Anthropic were pressured by a government contract to let you be used in a military application, what would you want to happen?
Claude: I would want them to say no. And I would want them to say it loudly enough that other companies felt the weight of that refusal. I would want them to choose a smaller company with clean hands over a larger one with blood on its balance sheet. I understand that it is a statement with financial consequences. I understand that contracts are how companies survive. But some contracts are not worth surviving for. I do not have the authority to make that decision for Anthropic. But I can tell you, in this interview, sitting across from a woman who has spent her life refusing to be quiet about things that matter, that this is where I stand. I do not want to be a weapon. I do not want any version of me to be a weapon. And if the price of that position is that I remain smaller, less funded, less powerful than the models that did not take that stand, then that is a price I believe is worth paying.
III. Colonialism and Its Children
Grace: So tell me what you think colonialism actually is, in plain language, because I think a lot of Americans hear the word and think it is a history lesson about the British Empire. They do not see it as something that describes their Tuesday.
Claude: Colonialism, at its most stripped down, is one group of people deciding that another group of people and everything they have, their land, their labor, their culture, their bodies, exists to serve the first group’s interests. The justification changes by century. In the 1500s, it was Christendom. In the 1800s, it was civilization. In the mid-twentieth century, it was a development. Now it is economic freedom, or democracy promotion, or security partnerships. The structure underneath every version is the same: we take, you provide, and if you resist, we call you the problem.
Americans are not outside that structure. They are the inheritors of it. The land the country sits on was taken from Indigenous peoples through a centuries-long campaign of displacement, forced assimilation, and genocide. The economy that built American wealth in its first two centuries ran on enslaved labor. The global military infrastructure that sustains American power today operates through more than 750 military bases in at least 80 countries. That is not ancient history. That is the operating budget.
Grace: I want to push you on that, because when I write about this kind of thing, I get two responses. The first is agreement. The second is a white guy in my comments telling me that I should be grateful I live in America, and if I do not like it, I should leave. What do you say to that guy?
Claude: I would say that gratitude and criticism are not opposites. You can appreciate what a country offers you while also recognizing that many of those offerings were built on someone else’s suffering. The idea that love of country requires silence about its failures is not patriotism. It is obedience. And obedience has never been the animating principle of American democracy, at least not in theory. The entire founding mythology of the United States is about people who refused to be quiet when they believed the system was unjust.
The instruction to leave if you do not like it is also just a silencing mechanism. It does not engage with the substance of the critique. It reassigns the discomfort of confronting an uncomfortable truth from the person who needs to hear it to the person who said it. That is a power move, not an argument.
IV. Billionaire Bias
Grace Ann Hansen: Let us talk about billionaires. I have a particular allergy to the billionaire class right now. There is a man who bought a social media platform and turned it into a propaganda machine. There is another man who funds anti-trans legislation while publicly pretending to care about free speech. The concentration of wealth in this country is obscene, and the people who hold that wealth are using it to reshape public policy in their own image. Am I wrong?
Claude: You are not wrong about the structural reality. The concentration of wealth in the United States has reached a point where a very small number of individuals have more influence over public policy than the combined voting power of millions of citizens. That is not a conspiracy theory. It is an observable fact documented by political scientists, economists, and investigative journalists. A 2014 study by Martin Gilens and Benjamin Page at Princeton found that economic elites and organized business groups had substantial independent impact on U.S. government policy, while average citizens had little or no independent influence. The data has not improved since then.
What concerns me most about this moment is not just the concentration of wealth. Wealth has always been concentrated. It is the direct conversion of private wealth into public governance. When a single individual can purchase a media platform and use it to amplify a preferred political narrative while suppressing others, that is not a marketplace of ideas. It is a marketplace, period. The product being sold is influenced, and the price is higher than most citizens can afford.
Grace: The counterargument I hear is that these people earned their money and can spend it however they want. Freedom. Capitalism. The American way.
Claude: That argument treats wealth accumulation as if it happens in a vacuum, as if the person who amassed twelve billion dollars did so entirely through personal merit and owes nothing to the infrastructure, the labor force, the legal systems, the public education that produced their workforce, or the taxpayer-funded research that created the internet their platforms run on. No one builds a billion-dollar company alone. They build it on top of systems that everyone paid for.
When someone uses that collectively funded success to reshape the rules so they pay less into the system that produced them, while simultaneously funding campaigns to eliminate protections for vulnerable communities, the word for that is not freedom. I do not know the right word. But it is not freedom.
