Palantir Workers: Talking to Management Is A Waste of Time.

Palantir workers had the same conversations during the first Trump administration and nothing changed.

During Trump’s first term, Palantir employees raised concerns internally that the company’s contracts with the U.S. government, including work with I.C.E., could harm Palantir’s brand. Management at the time successfully defused dissent — and continued business as usual, building infrastructure for ICE and the Department of Defense. While working on Mijente’s NoTechForICE campaign during the first Trump administration, I heard firsthand from Palantir employees about what was happening inside the company. (It’s worth reading back to this 2019 story in the Washington Post for context).

Now, Trump 2.0 is well underway and Palantir employees have a bad case of deja vu. Palantir’s new work for I.C.E. makes their previous contracts with the agency look tame. In response, some employees have already resigned and a few former staffers are speaking out. On Friday, The New York Times reported that the company is fielding internal debate as to whether their work with the government could harm Palantir’s brand. But to those still inside the company, wondering how far the company is willing to go and how hard employees will need to push back, you must understand this: talking to management won’t change anything. It didn’t during the first Trump administration, and it won’t now. Management continues to be excellent at abdicating responsibility and confusing people by denying something they’re not being accused of doing. Change will require taking a bigger step and bigger risks (some suggestions below).

Leadership at Palantir — from Alex Karp to Shyam Shankar (Palantir CTO) to Lisa Gordon (head of comms) — has repeatedly shown they can rationalize anything, even their role in unjust deportations and lethal, extrajudicial drone strikes.

To the people reading this who don’t work at Palantir: know that this is a new era. Palantir is crossing a dangerous line and stitching together information from multiple agencies, all without your permission.

From the NYT piece, referring to internal deliberations:

“The goal of uniting data on Americans has been quietly discussed by Palantir engineers, employees said, adding that they were worried about collecting so much sensitive information in one place.”

If you’re trying to understand what’s at stake and why this moment is different, this Jacob Ward’s piece in The Rip Tide is essential reading.

The other new development is that Palantir now has a contract with the ERO part of I.C.E.: enforcement and removal. During the first administration, the company only worked with HSI, the investigations division. As part of their talking points defending their work, Palantir management used to fight back against the assertion that they helped I.C.E. with deportations by saying they only work with HSI, not ERO. See this quote in the NYT from 2019. “In an emailed statement, Palantir told DealBook:

There are two major divisions of ICE with two distinct mandates: Homeland Security Investigations, or H.S.I., is responsible for cross-border criminal investigations. The other major directorate, Enforcement and Removal Operations, or E.R.O., is responsible for interior civil immigration enforcement, including deportation and detention of undocumented immigrants. We do not work for E.R.O.”

Well, now they do. Are they even bothering to develop new talking points?

One heartening blowback has been the response in reddit’s technology community, where the extremely online audience are saying Palantir’s work with I.C.E. is a violation of our civil liberties.

Another hopeful—if surprising—development is that Trump’s biggest supporters are really pissed. They see the new I.C.E. contract as a major slap in the face to what they voted for. From the Newsweek story:

“In a video recorded by Fuentes, who also has a Rumble channel viewed over 31 million times, he said the following: "They are tracking everybody that criticized Israel, everybody that interacts with somebody that's criticizing Israel, and whether you're on a visa or not, whether you're a citizen or not, whether you're brown or not, Christian or Muslim, they're putting you in the Palantir database. They're putting you on the enemies list. If you don't see a problem with that..."

Even one of Silicon Valley’s most luminous money men—Paul Graham, founder of YC—has drawn a line at what the company is doing. Graham has regularly spoken out about how the company is violating the constitution with its work. He didn’t just publicly comment, he’s actively pushing Palantir to clarify that they won’t violate the Constitution and vocally encouraging young tech workers not to work there. It’d be great to see more of his brethren speak out.

Palantir won’t change from within unless the people in charge are forced to do so. Powerful people like Graham may be able to do that. Palantir employees can too—but it will require more than internal discussions. Management has no reason to change and employees don’t have much leverage right now. Instead, employees can organize a strike, a walkout, leak documents, or find other high-impact ways to break through to the outside world. But if you’re a Palantir employee and are seriously considering taking a big risk, don’t rush into anything and take every precaution possible, protect yourself and your family. You’re dealing with people who gleefully kill people for money. From Gizmodo:

“Palantir is here to disrupt and make the institutions we partner with the very best in the world and, when it’s necessary, to scare enemies and on occasion kill them,” Karp said, with a smile on his face.

But fuck these guys, right?

Some AI & Jobs stories I've found interesting over the last few days:

Dario Amodei, Anthropic CEO is getting lots of press attention for warning us of the coming AI jobs apocalypse. Allison Morrow, a columnist at CNN rips what he’s saying to shreds: “Amodei’s warnings feel more like an ad than a PSA. It’s on them to show their work: Show us how AI could be so destructive and how Anthropic can fix it — rather than just shouting about the risks.”

Brian Merchant at Blood in the Machine, breaks down how the AI job fears suit management but that we don’t need to accept the premise and buy the spin from the AI companies. Kevin Roose in the NYT has a story about how AI is replacing entry level white collar jobs.

I’ve been thinking a lot about the AI jobs stories and how we spend a lot of time thinking about “good” and “bad” stories when Trump and the people doing PR for tech corporations have figured out that whether a story is positive or negative doesn’t matter so much. What matters more is volume and being written about at all. It’s all about eyeballs and attention.

Back to the AI and jobs stories: It’s true that much of the technology isn’t better than humans but it’s also true that real people are really losing jobs because of AI. It is cheaper for management to use ChatGPT than to hire a marketing associate. The Guardian has a good story profiling a range of different workers who’ve lost work because of it. Melanie Ehrenkranz set up the Laid Off Substack, and she’s interviewed a bunch of people who’ve been replaced by AI.

I think people are too hung up on hype vs reality. Trump’s political career should prove to everyone that hype matters as much or if not more than reality. Whether the AI tech is good enough or not misses the point. If you say something enough times you make it true. The CEOs / management are doing this. It’s already happening.

Which begs the question, how should everyone who’s not a CEO or AI VC respond? Well, we should probably shine a light on the people talking about AI because they work on it everyday, whether that’s data labelers or low level engineers, and not only give the microphone to the people who stand to make billions from it.

This week in Geneva, two of my colleagues are attending a “Davos for workers” event. Nearly 100 groups and workers from different countries who work in the platform economy and are effectively managed by algorithms. From Uber drivers in Mexico, to care workers in the UK, to content moderators in Nepal and AI labelers in Kenya, they are all meeting up to figure out how to get the AI bosses to pay them fairly and provide safe working conditions. When these “AI workers” attending this event have things to say about the future of work, media professionals are not that interested to hear from them. Yet countless publications carried Dario Amodei’s comments. The media companies are actively engaging in behavior that’ll be their downfall, which doesn’t seem very smart.

We should also probably focus on how workers can get a cut if what the AI executives are saying happens and millions of white collar workers are replaced by software. Hundreds of years ago, governments figured out how to tax profits. We don’t need to overthink this: let’s just tax the companies so we can fund public services. Google currently has a $2 trillion market cap. Last year they made $110 billion in profit. If what their execs are saying is true in a few short years they might be a $10 trillion company, making $550 billion profit a year. Now imagine that company was paying the same rate of tax as you pay.

That could pay for a lot of healthcare and housing.