Everyone loves a morality play. The media looks at Palantir posting blowout earnings while facing relentless criticism over its military contracts in Gaza and Ukraine, and they immediately try to fit the story into a tired script. The lazy narrative writes itself: evil corporation profits from human suffering, stock goes up because conscience is expensive.
It is a clean story. It is also completely wrong.
I have watched enterprises blow millions trying to build custom software in-house while missing the actual mechanics of why government and defense procurement works. The financial acceleration happening right now inside Alex Karp’s company is not a referendum on corporate ethics. It has nothing to do with whether boardrooms approve or disapprove of foreign policy.
Palantir is growing because enterprise software is broken, legacy IT is a money pit, and nobody else can actually connect disparate databases without setting fire to the organization chart.
The Moral Panic Fallacy
Let us address the elephant in the room immediately. Critics want you to believe that Palantir is an outlier, a uniquely controversial entity whose financial results are weighed down or somehow miraculously propelled by geopolitical friction. This is naive.
Defense budgets do not move because Twitter users are upset about military logistics software. Governments buy platforms because they have petabytes of unstructured data and zero internal capability to make sense of it. When a state agency needs to track supply chains, parse intelligence reports, or manage operations across disparate branches, they do not care about op-eds in major newspapers. They care about survival and administrative competence.
For years, I have listened to tech commentators argue that public backlash will force institutional clients to abandon efficient tools for softer alternatives. It never happens. When an organization is drowning in bureaucratic inefficiency, moral branding takes a back seat to operational survival. Palantir understands this better than any Silicon Valley startup that spends its time worrying about culture fit and carbon offsets.
The growth is otherworldly, sure, but not because of dark magic or controversial battlefield deployments. It is happening because the commercial market finally realized that Palantir’s ontology-based approach actually works, while traditional cloud dashboards are expensive digital wallpaper.
The Myth of the Custom Build
To understand why Gotham and Foundry are eating the market, you have to look at what companies were doing before they surrendered to software licensing fees.
For the last decade, CIOs bought into a dangerous myth: that you could build your own data infrastructure using open-source tools, a dozen microservices, and a massive army of mid-level engineers. I have seen Fortune 500 companies burn fifty million dollars trying to stitch together data lakes that turned into bottomless swamps.
Imagine a scenario where an enterprise has thirty legacy databases, ranging from mainframes installed during the Reagan administration to modern cloud apps running on AWS. The standard advice from tech consultants is to hire a hundred engineers to build custom pipelines. Two years and a hundred million dollars later, the pipelines break every time a database schema changes by a single character.
Palantir’s core innovation, Foundry, bypasses this entire nightmare. It creates an operational ontology—a digital twin of the business that maps how data points relate to one another in the real world. It does not ask you to migrate your data. It does not force you to rewrite your legacy code. It sits on top of the mess, makes sense of it, and gives operators a UI that actually lets them make decisions.
When a product saves a company hundreds of hours of manual engineering grunt work every single week, nobody cares about the corporate PR cycle. They renew the contract.
Why the Commercial Segment is Winning
The lazy narrative says Palantir is just a defense contractor with a Silicon Valley aesthetic. The numbers prove otherwise. The commercial expansion, particularly in the United States, is where the real story lies.
For a long time, Wall Street dismissed the company as a consulting shop disguised as a product company. They pointed to heavy reliance on government contracts and argued that commercial clients would never adopt such an intrusive, top-down platform. They missed the shift entirely.
Enterprises are desperate. They spent the 2010s collecting data like hoarders, and now they are drowning in it. They have data warehouses full of metrics, but zero operational velocity.
Palantir introduced the Artificial Intelligence Platform, or AIP, not by running splashy consumer ad campaigns, but by running bootcamps. This was a masterclass in enterprise sales execution. Instead of pitching a slide deck about the theoretical future of machine learning, they locked enterprise executives in a room for a few days, hooked AIP up to their actual messy data, and let them build working LLM-driven operational workflows before lunch.
When you show a supply chain director an interface where they can type a prompt and instantly reroute cargo to avoid a port strike, you have won the deal. You do not need a moral argument. You have operational proof.
The Downside Nobody Wants to Admit
Let us be completely candid about the risks, because contrarian analysis means calling out your own thesis when it wobbles. Palantir’s approach has a massive, glaring downside: extreme vendor lock-in.
When you build your entire operational backbone on top of Foundry and AIP, you are essentially adopting a digital landlord. You do not own the ontology. If your relationship with Palantir sours, or if pricing models shift aggressively in the future, untangling yourself from their ecosystem is nearly impossible. It is the architectural equivalent of marrying someone who controls all the oxygen in the room.
Traditional tech companies preach open standards and interoperability while delivering fragmented junk. Palantir preaches proprietary dominance and delivers absolute control. For many enterprises, the control is worth the price of admission. For others, it is a ticking time bomb of future debt.
Wall Street analysts who obsess over quarterly political headlines are completely missing this structural reality. The real risk for Palantir shareholders isn't bad press; it is whether the company can scale its deployment model without relying on an army of deployed engineers to hold customer hands through every configuration.
The Real Question We Should Be Asking
People keep asking whether companies should work with a software vendor that takes controversial defense contracts. That is a comforting question because it frames business success as a moral choice for consumers and investors to judge from afar.
It is also the wrong question.
The question you should be asking is why every other major enterprise software player failed to build an operating system that actually matters to the core workflow of an enterprise. Why are traditional software giants still selling systems of record—glorified filing cabinets—while Palantir built a system of action?
Until SAP, Salesforce, or Oracle figure out how to give enterprises an operational ontology that replaces the need for custom engineering armies, Palantir will keep printing money.
The protests will fade. The headlines will change. The balance sheet will remain. Software that works beats software that makes you feel good every single time.
