Why Palantir Employees Are Revolting Against ICE Contracts

After federal agents shot and killed Minneapolis nurse Alex Pretti on Saturday, something unprecedented happened inside Palantir's corporate walls: employees started openly questioning whether their company should be helping ICE at all.

Internal Slack messages reviewed by WIRED reveal a workforce in ethical turmoil. "In my opinion ICE are the bad guys," one employee wrote. "I am not proud that the company I enjoy so much working for is part of this."

The rare public dissent highlights a growing tension between Palantir's lucrative government contracts and its employees' moral compass—a conflict that's becoming harder to ignore as immigration enforcement intensifies under the Trump administration.

The $30 Million Question

Palantir's relationship with ICE centers on a $30 million contract signed in April for a platform called ImmigrationOS. The system provides three core functions:

• Enforcement Operations Prioritization and Targeting
• Self-Deportation Tracking
• Immigration Lifecycle Operations focused on logistics planning

According to the company's internal wiki, the platform gives ICE "near real-time visibility" into people self-deporting and helps agents identify who to deport. The pilot was renewed in September for another six months, and the company recently started a new project with USCIS to identify "fraudulent benefit submissions."

When Employees Demand Answers

As worker questions flooded company Slack channels, Palantir's privacy and civil liberties team scrambled to respond. Courtney Bowman, the company's global director of privacy and civil liberties engineering, linked to an internal wiki defending the work as improving "ICE's operational effectiveness."

But employees weren't satisfied with corporate talking points. When one worker asked whether ICE could use Palantir's platform beyond contracted scope—like pulling data from outside sources—CTO Akash Jain's response was blunt: "Yes, we do not take the position of policing the use of our platform for every workflow."

That admission is significant. It means ICE could potentially expand DHS's surveillance capabilities by integrating third-party commercial data or information from other agencies—exactly what civil liberties advocates fear most.

The Database Nobody Talks About

The internal debate intensified when an employee shared a video showing an ICE agent telling a legal observer: "We have a nice little database, and now you're considered a domestic terrorist."

When asked if Palantir provided such a database, Jain claimed he wasn't "tracking any database like this." But the company declined to comment publicly on whether its software powers surveillance tools used against legal observers and activists.

The wiki acknowledges "increasing reporting around U.S. Citizens being swept up in enforcement action" but argues that Palantir's ICE customers "remain committed to avoiding the unlawful/unnecessary targeting" of citizens. It's a careful legal distinction that may not satisfy employees watching enforcement operations on the news.

The Business vs. Ethics Calculation

Palantir's federal contracts have exploded to over $900 million during Trump's first year in office. Beyond ICE work, the company is building a "mega API" for the IRS and maintaining extensive military contracts.

One employee captured the tension perfectly: "Thinking pragmatically: is the reputational damage we're taking for being associated with them worth it? What if the next administration will be democratic and they cut all the contracts with us?"

It's a question that cuts to the heart of Palantir's business model. The company has built its success on government contracts that often involve surveillance and enforcement. But as those operations become more visible and controversial, the human cost becomes harder to ignore.

When Technology Meets Accountability

Palantir's defense relies on a familiar tech industry argument: we build tools, others decide how to use them. The company says it implements "strong controls" but acknowledges "there won't be bad apples, mistakes or other issues that lead to adverse outcomes."

It's a position that increasingly rings hollow to employees watching their technology deployed in operations that separate families and target communities. The company may not "police" every workflow, but its software makes those workflows possible at unprecedented scale and precision.