Alex Karp and the education of an argument

In Frankfurt, before the government contracts and the earnings calls, Alex Karp turned up at an academic’s office with a backpack. Karola Brede remembered a young researcher trying to find his place in German university life. He presented his qualifications. She suggested a tutoring position and a working group at the Freud Institute. The future chief executive of Palantir began, in this recollection, with the familiar business of finding somebody who would take his work seriously.

There is a temptation to read the scene backward: the unconventional visitor was already a technology founder in disguise. Brede’s memory resists that convenient plot. She saw someone seeking an academic career. The remembered scene belongs to an academic career, with no business launch in view. The distance between that office and Palantir’s boardroom matters because Karp has spent much of his public life arguing about institutions, while acquiring the power to supply them.

A backpack in Frankfurt

Born in New York in 1967 and raised around Philadelphia, Karp grew up with parents involved in civil-rights activism. He attended Central High School, studied philosophy at Haverford, and graduated in 1989. Stanford Law followed. There he met Peter Thiel, a relationship that would outlast their time in the classroom and eventually become a business partnership.

His next destination was Goethe University Frankfurt. He completed his doctorate in 2002, writing in German about aggression, jargon, and culture. His dissertation drew on social theory, including Talcott Parsons and Jürgen Habermas. Brede supervised the work. Habermas’s influence was real, but the distinction between an influential thinker and a dissertation supervisor deserves to survive the trip into a short biography.

The subject also deserves better than being filed under “philosophy.” It concerned how aggression could fit inside a society’s accepted practices rather than appear only as disruptive behavior outside them. That question has an uncomfortable range. An institution can maintain order and still normalize things that deserve scrutiny. Reading the later career alongside the early work invites questions; it does not establish that a dissertation secretly designed a company.

The friendship that became a company

Palantir was founded in 2003. Thiel’s experience at PayPal helped supply an initial idea: methods for detecting fraud might also help intelligence organizations find adversaries. The founding group included engineers as well as Karp. Early backing came from In-Q-Tel, the CIA-linked investment organization. Karp’s position at the company would put him between technical work, customers, and the political meaning of both.

A decade later, Palantir was still a private business whose work could be difficult for outsiders to inspect. By early 2015, it had attracted more than $1 billion in investment. Government agencies, banks, and corporations had become part of its customer world. An enterprise software business had grown around work that was often sensitive, while its chief executive remained a person trained to examine the assumptions behind an institution.

Karp tells the founding story with less ceremony. Recalling Thiel’s invitation at Davos in 2023, he described accepting without first negotiating the compensation details. In the same conversation he remembered telling German colleagues that an excessively strong shampoo had produced his hair, and that a lawsuit explained how he could afford to live there. The story was a joke. Evidently, American litigation was plausible enough to make it travel.

The humor helps explain his interview presence. Karp can move from an absurd personal anecdote to a dispute about national security without giving the audience much time to put down its coffee. The hair gets attention. The propositions underneath it require rather more effort.

Five words before the listing

In 2020, as Palantir prepared to enter the public markets through a direct listing, Karp’s founder letter stated its position in five words: “We have chosen sides.” The letter linked the company’s work to the United States and its allies. It also criticized a technology industry that could profit from personal data while hesitating to support public institutions.

“We have chosen sides.”

Alex Karp · founder letter, 2020

A listing document is usually where conviction gets dressed for a meeting with accountants. Karp gave investors an argument about allegiance. The declaration made his preferences legible, and made it harder to treat Palantir’s customer choices as routine procurement. He was asking the market to buy into a company whose chief executive intended to explain its work in political terms.

His 2022 annual letter returned to the relationship between the governing elite and the military. It disputed the assumption that difficult moral and political questions had already been settled. The recurring concern was institutional distance: people dependent on national security could nevertheless regard the people and technology providing it with suspicion.

That framing also leaves a demanding question for Karp. If engineers should engage seriously with public power, then objections to the uses of their software belong inside the conversation. Choosing a side establishes an orientation. It cannot, by itself, settle each decision made under that banner.

When the argument went to Kyiv

On June 2, 2022, Karp met Volodymyr Zelenskyy and Ukraine’s digital transformation minister, Mykhailo Fedorov, in Kyiv. The meeting concerned defense, security, and digitization. Russia’s full-scale invasion had begun a few months earlier. A chief executive whose letters discussed Western institutions was now visiting a government fighting to preserve its own.

Ukraine’s president welcomed the visit as a signal that the country remained open to cooperation. Karp said Palantir was ready to establish an office and work with Ukrainian specialists. The encounter gave a concrete setting to his preferred account of technology: software could belong to a national effort, rather than remain a service bought at a distance.

The photograph from the following year’s AI summit at Bletchley Park shows another part of that role. Karp stands beside Britain’s then deputy prime minister, Oliver Dowden. Governments had become places where he could be received, heard, and photographed. The backpack had acquired a conference badge.

Government access also brings scrutiny. Civil-liberties groups have challenged Palantir’s immigration work, including ImmigrationOS. In 2025, Amnesty International argued that such tools presented surveillance and human-rights risks. Palantir responded that its ICE work did not target student protesters. Those are competing positions about consequential capabilities, with people beyond the contracting parties at stake.

Karp has offered his own security argument: effective government action against threats can help prevent a later, broader assault on civil liberties. The claim deserves examination alongside the critics’ concerns. One asks what happens if governments lack useful tools; the other asks what happens when governments possess them. His career puts both questions in the same room.

A republic between hard covers

In February 2025, Karp and Nicholas Zamiska published The Technological Republic: Hard Power, Soft Belief, and the Future of the West. The book asks the software industry to recover a sense of public purpose. Its central grievance is that an extraordinary concentration of engineering talent has been directed toward consumer convenience and advertising while urgent national challenges have received too little attention.

The prescription includes a renewed relationship between government and technology companies. It also asks leaders to tolerate ideological confrontation. In that respect, a book gave Karp more space for a familiar concern: a culture too anxious about disapproval may struggle to produce either serious beliefs or ambitious work.

There is an appealing difficulty in asking successful engineers to think beyond the next application. There is another difficulty in deciding who names the common purpose they should serve. National ambition is easy to recommend in the abstract. Particular policies, customers, and weapons require particular judgments.

An existing institution complicates that prescription: DARPA already connects government-supported research with technical innovation. Its limited place in the book leaves a useful question about how much of the desired relationship needs rebuilding, and how much needs recognizing. A proposal to renew cooperation has to account for the cooperation already there.

Karp’s aspirations have therefore become unusually available for inspection. Readers can compare a published political project with the company he runs. A book can make an argument coherent. The contracts will keep supplying complications.

A cabin-sized decision

In August 2021, Karp wrote a personal check for $180,000 to David Lidstone, known in New Hampshire as River Dave. Lidstone had spent nearly three decades living on land he did not own. A property dispute led to his imprisonment, and his cabin burned down. The donation was intended to help him build a home elsewhere.

A lawyer delivered the check. A company spokesperson confirmed the gift without elaborating. This is a smaller scene than Kyiv or a public listing, and it should remain one. It does not resolve an argument about software or establish a theory of the donor’s character. It does show Karp making a specific personal intervention when the problem was somebody’s shelter.

Public profiles can make every anecdote audition for the job of explaining a whole person. The cabin story is useful precisely because its scale is limited. There was a recipient, a loss, a sum, and a stated use for the money. For once, the reader can set down the geopolitical map.

The argument now comes with a revenue chart

By August 2026, Karp’s public argument had acquired another target: the way enterprise AI was being sold. In interviews and his second-quarter shareholder letter, he questioned whether purchasing model usage necessarily produced useful business results. He also warned about companies surrendering data and organizational knowledge to the suppliers of language models.

The distinction is practical. A customer can consume an impressive amount of computation without improving a decision or producing something worth the bill. Karp presents Palantir as a business aligned with the value its customers create. That is his commercial claim, and a claim customers can test against their own experience.

Palantir reported $1.94 billion in second-quarter revenue, up 93% from a year earlier. Its U.S. commercial business generated $764 million, while U.S. government revenue reached $809 million. The figures show that the company’s growth reaches across two audiences. They also mean more people have a reason to listen when its chief executive speaks.

Commercial expansion gives a worldview a larger platform. It supplies no automatic verdict on that worldview. Karp’s original questions about institutions have followed him into a position where the answers can affect customers, employees, citizens, and soldiers. The young researcher needed a place for his work. The executive has one. What happens inside it remains open to argument.