The greeting wasn’t “hello.” When two former Palantir employees reconnected by phone late last year, the first words on the line were: “Are you tracking Palantir’s descent into fascism?”

“That was their greeting,” the second former employee told Ars Technica. “There’s this feeling not of ‘Oh, this is unpopular and hard,’ but ‘This feels wrong.’”

Inside Palantir — the Peter Thiel-cofounded data analytics giant born from CIA seed money after September 11 — a confrontation is accelerating. The company’s workforce is increasingly at odds with its expanding role in the Trump administration’s immigration crackdown and military campaigns, and the people sounding the alarm are the ones building the tools.

A Mission That Moved Inward

Palantir was founded at a moment of national consensus. After 9/11, building powerful surveillance software to hunt terrorists abroad was something many in Silicon Valley could rationalize — even if the company’s namesake, Tolkien’s all-seeing and corrupting orb, suggested the founders understood the risks from the start.

For two decades, employees absorbed external criticism and awkward dinner-table conversations about working for a surveillance company. Palantir sold its data aggregation and analysis platforms to private businesses and the US military’s targeting systems, and workers could tell themselves the mission was defensible. The enemy was foreign. The targets were terrorists.

That calculus shifted last fall. According to Ars Technica, Palantir seemed to become the technological backbone of Trump’s immigration enforcement machinery, providing software to the Department of Homeland Security that identifies, tracks, and helps deport immigrants. A year into Trump’s second term — with the administration’s escalated immigration enforcement, its war in Iran, and Palantir’s own public manifestos — employees are raising concerns internally with new urgency, according to current and former workers who spoke with the publication.

“Fierce Internal Dialogue”

A Palantir spokesperson defended the company’s culture in a statement to Ars Technica. “We hire the best and brightest talent to help defend America and its allies and to build and deploy our software to help governments and businesses around the world. Palantir is no monolith of belief, nor should we be. We all pride ourselves on a culture of fierce internal dialogue and even disagreement over the complex areas we work on. That has been true from our founding and remains true today.”

The statement is both a defense and an inadvertent confirmation. A company whose employees are debating whether its work constitutes a “descent into fascism” is one where the mission is no longer settled — and a spokesperson insisting that internal disagreement has existed “from our founding” doesn’t answer the obvious question: what changed?

What changed is the direction the tools are pointing. Palantir’s government work has moved from counterterrorism — tracking suspected militants abroad — to domestic immigration enforcement, identifying and facilitating the deportation of people inside the United States. For employees who signed up to build systems targeting foreign threats, the domestic application has become a breaking point.

The Pipeline Problem

Palantir is not the first tech company to face employee revolt over government contracts. Google workers protested Project Maven, the Pentagon’s AI-targeting program, in 2018. Microsoft employees objected to ICE contracts. Amazon staff pushed back on facial recognition sales to police departments. Each time, the pattern was similar: workers who built the tools discovered those tools were being used in ways they hadn’t anticipated — or had been promised wouldn’t happen.

But Palantir occupies a different position. This is not a company with a consumer brand to protect or a diversified revenue base that makes government contracts optional. Government work is the business model. There is no other line of business to fall back on, no advertising revenue to cushion a moral stand.

The employees raising concerns are not external critics. They are the people with the institutional knowledge to understand exactly how these systems work — and the technical skill to imagine alternatives. Their alarm carries weight that outside commentary cannot replicate.

This is the broader test case that Palantir now represents: whether tech workers, from inside the building, can meaningfully slow or redirect the authoritarian applications of their labor once the contracts are signed and the infrastructure is running. Whether “fierce internal dialogue” produces anything more than dialogue.

The company’s namesake, Tolkien’s palantír, was a seeing stone that corrupted those who used it. The point was not that surveillance itself was evil — it was that the power to watch from a distance corroded even those who started with good intentions.

Palantir’s employees appear to have read the book. The question is whether anyone with the power to change course is listening.

Sources