This podcast is brought to you by Progressive Insurance. Fiscally responsible, financial geniuses, monetary magicians. These are things people say about drivers who switch their car insurance to Progressive and save hundreds. Because Progressive offers discounts for paying in full, owning a home, and more. Plus, you can count on their great customer service to help you when you need it, so your dollar goes a long way. Progressive Insurance.

Visit Progressive.com to see if you could save on car insurance. Progressive Casualty Insurance Company and affiliates. Potential savings will vary, not available in all states or situations. This message is brought to you by Apple Pay. Forget your wallet, it's all good. Because with Apple Pay, you can pay with a simple tap of your iPhone, the wallet you never forget, at millions of places worldwide, including websites, apps, and anywhere you see the contactless symbol.

Security is built in with Face ID, so you don't have to worry about your cards getting lost or stolen. And the best part, you still earn the card rewards, points, and cash back you love. So say goodbye to the buy fold. Add your card to Apple Wallet and start paying the Apple way. Terms apply. Back in April, I read a New York Times story that stuck with me. It was about all the data the federal government might have about us.

The government is going to know everything from the expected, like your social security number and your home address if you've bought a house and filed taxes, to more unusual things. If you are a vet, they might know a lot about your mental history and your personal family history. That's Shira Frankel, one of the reporters who wrote the story.

She and her co-author Emily Badger found 314 pieces of data that the government might have about individual Americans.

If you are a student who's ever applied for a student loan, they might know a lot about your extended family, your cousins, your aunts, your uncles, where they went to school, if they're U.S. immigrants, if they live in other countries. It's amazing once you think about all the different government agencies that we give data to, how much of our personal lives we kind of let them know about.

Shira wrote that story after the Trump administration issued an executive order calling for the consolidation of government records. At the time, one of the biggest questions was what the administration wanted from that consolidation. Or to put it another way, where was all that data going? Now, thanks to Shira's reporting, we have an answer.

So at least one place that we know that it's going or at least being organized is a company called Palantir. One of the largest data analytics companies in the country. One with deep ties to the federal government and to Elon Musk's so-called Department of Government Efficiency.

And so what we've learned is that the Doge team has been really aggressive in pushing Palantir into new contracts in the government because they see it as this really sort of smart, clean solution for them of being one place, one company that can organize all this data across different government systems. And then ultimately, if they so choose to, co-operate.

compile it, collate it into one place. And that, what I mean by that is taking data that you may have sent the IRS, taking data you may have sent, you know, the Department of Education for loans, the Department of Homeland Security because of a welfare program, taking all those different data points and putting it in one central place.

What you're describing sounds like a centralized data lake. Is that accurate? Yeah, that's a good way of thinking about it. And I will note it's something that other countries have in place. We know that countries like Russia and China do this with their citizens. They centralize data in one place on all of their citizens.

We are not Russia or China. We're not. And that's an argument that has been made for decades by privacy advocates and civil liberties groups who say there's a lot of reasons we don't want to be Russia and China. But one is that people trust us with their data for different reasons. And when you organize it all in one place, it gives you the power to do a lot. It gives you the power to know a lot.

And we're worried what a government will do with this power, depending on administration on a particular White House and its goals. Today on the show, how Palantir is laying the groundwork for a central government data system. One that sounds an awful lot like a surveillance state. I'm Lizzie O'Leary, and you're listening to What Next TBD, a show about technology, power, and how the future will be determined. Stick around.

This message is brought to you by Apple Pay. Forget your wallet, it's all good. Because with Apple Pay, you can pay with a simple tap of your iPhone, the wallet you never forget, at millions of places worldwide, including websites, apps, and anywhere you see the contactless symbol.

Security is built in with Face ID, so you don't have to worry about your cards getting lost or stolen. And the best part, you still earn the card rewards, points, and cash back you love. So say goodbye to the buy fold. Add your card to Apple Wallet and start paying the Apple way. Terms apply.

This episode is brought to you by Agency. While single agents can handle specific tasks, the real power comes when specialized agents collaborate to solve complex problems. But there's a fundamental gap. There's no standardized infrastructure for these agents to discover, communicate with, and work alongside each other.

That's where Agency comes in. The Agency is an open-source collective building the Internet of Agents, a global collaboration layer where AI agents can work together. It will connect systems across vendors and frameworks, solving the biggest problems of discoverability, interoperability, and scalability for enterprises.

with contributors like Cisco, Crew AI, Langchain, and MongoDB, Agency is breaking down silos and building the future of interoperable AI. Shape the future of enterprise innovation. Visit agency.org to explore use cases now. That's A-G-N-T-C-Y dot O-R-G.

Let's back up and talk a little bit about Palantir generally. How did it start? Who runs it? What do they do? So Palantir started over a decade ago. They kind of looked at data and data systems that were out there and thought there's a better way. They were the first really to go over the government aspects.

As a client, they actually had to first sue the Department of Defense in order to get their first military contracts. And they went in and they said, we can do this better. We can give you an off-the-shelf software tool that will help you not just organize your data, but visualize it.

And by all accounts, they were good at this. They worked with the FBI. They worked with police departments all over the United States. They worked with the Department of Homeland Security in helping catch alleged terrorists. They were really good at looking at data. And then even for people who had no kind of technical background...

they could present it in a way which was like, hey, here's the conclusions drawn from this data. You were looking for an individual with ties to a militant group. Here is where we think that individual is based on all these different data points that you've inputted into our system. Were they accurate?

By all accounts, they were. And I will note that both Democratic and Republican White Houses have brought their technology into the government. They were brought in by the Biden administration to help the CDC organize the distribution of vaccines across the United States during COVID. So we've seen different, you know, Democratic leaders and Republican Party leaders say that they like their software and they want to use it.

Palantir was co-founded by Peter Thiel. And of course, avid readers of Tolkien will note that a Palantir, the Palantiri are like these magical orb things in Lord of the Rings. But Thiel is not currently running the company. Alex Karp, the CEO, is. Tell me a little bit about the players behind this company.

Well, you know, Alex Karp has come out publicly talking about how he's a Democrat, but also talking about how he admires parts of the Trump administration's agenda. You know, I just heard him speak the other month, actually, in Washington, D.C., and I was struck by his kind of political pragmatism.

in his want to kind of appeal to people across the political spectrum by saying that ultimately for him, and I think for what his company represents, is that this technology is agnostic and how it's used is kind of determined by the person operating it. And their kind of point of view is that like this data is incredibly powerful. This data is incredibly useful. Data is one of the most important kind of

currencies in modern society, and they want to be the ones that are organizing it efficiently. Palantir has had associations with all kinds of political administrations, but do you think it's fair to say the number of contracts or the reach of the contracts has amped up in the second Trump administration?

Oh, absolutely. I mean, we have found that so far from what we can confirm, and this is just what appears in public records, and there's a lag here, so we don't actually know the total amount, but there's been at least $113 million in funds that have been given and almost a billion that was allocated by the Air Force. And so we're seeing this huge ramp up in contracts with Palantir. And I think more importantly is kind of what we heard from people within government, which is that

If this continues, if this ultimate vision of Trump's continues, that this data will all be organized and centralized and all the silos will get removed, that Palantir will end up being an absolutely vital tool, not just within the Trump administration, but going forward, right? As the place to organize data in the U.S. government. That is an incredibly large and presumably an incredibly lucrative contract if that were to come.

If the plan is to compile unclassified data across different agencies, what are they trying to do with it? Just to get to that central repository point or do something with that? Because after all, it's a lot of work. It costs a lot to centralize all of that information.

Totally. And that's a great question. What do they plan to do once they centralize all this data? So far, we know what they've said, which is that they want to eliminate waste. They think it can help them look for fraud. I will note that people in the past have talked about how it could be really, really useful to centralize data if you want to make welfare more efficient in the United States, if you want to create a better social security net in the United States. There's lots of things you can do when you centralize data. What

We have also heard in our reporting over the last couple months that, for instance, if the Trump administration wants to achieve its goals of removing undocumented people from the United States, one way that you could do that and make it faster, and I suppose they would say more efficient to remove people, is by having all this data centralized. For years, people who are undocumented in the United States have been told to file taxes.

that they should still pay taxes. You know, you're living here, you're not documented, but pay your taxes. And millions of them have, right? Tons of them have paid taxes. And so if you were able to look through the IRS data and see, you know, who was filing as a non-citizen, I mean, that would really quickly give you an updated list of these people, where they live, where they go to school, where they work. A lot of information is in your tax return.

And traditionally, agencies have said, no, no, this is our fiefdom. You don't get to look at that other agency. You don't get to look at that for removal purposes, for example.

Right. So, for example, in that specific case of removal purposes, that data has been very protected. They've said, hey, we told people to give us this information for tax purposes. It's actually going against all these, the Privacy Act is going against the terms of our own contract to now use that same data towards another purpose. We told people this was safe. We told people to do this. We can't now go and use it to forcibly remove them from the United States.

When we come back, why some Palantir employees are sounding the alarm.

If you're running a business, you know how important it is to stay connected to your customers and having a flexible and efficient phone system is essential to succeed. Introducing OpenPhone, the number one business phone system that streamlines and scales your customer communications. OpenPhone works through an app on your phone or computer, and your team can share one number and collaborate on customer calls and texts like a shared inbox.

That way, any teammate can pick right up where the last person left off, keeping response times faster than ever. Plus, with AI-powered call transcripts and summaries, you'll be able to automate follow-ups, ensuring you'll never miss a customer interaction again. So whether you're a one-person operation or have a large team that needs better collaboration tools, check out Open Phone. See why over 50,000 businesses trust Open Phone to manage their businesses' calls and texts.

This podcast is brought to you by Progressive Insurance.

The

The Hoover Dam wasn't built in a day. And the GMC Sierra lineup wasn't built overnight. Like every American achievement, building the Sierra 1500 heavy-duty and EV was the result of dedication. A dedication to mastering the art of engineering. That's what this country has done for 250 years.

and what GMC has done for over 100. We are professional grade. Visit GMC.com to learn more. Assembled in Flint and Hamtramck, Michigan and Fort Wayne, Indiana of U.S. and globally sourced parts.

What steps do we know the administration has taken thus far to move forward with this project? We already know that they've brought Palantir into different government agencies where they have not formally had work or done this kind of work. They've expanded existing contracts and they've talked about bringing them into places like the Social Security Agency, which has not yet happened, but we know those talks are ongoing.

We also know on a much more basic level, they've started to merge some of this data. So some of my colleagues reported a few months ago that already people from the Department of Homeland Security have reached out to people from the Social Security Agency and said, hey, we're looking for these individuals. We're going to give you their Social Security numbers. Help us find them and track them down. So already...

already there's been a breakdown of some of those walls, some of those silos that had been intentionally created. Already we're seeing some of that happen. And I think what's so interesting about this is that anyone who's worked with data will tell you, once you merge data, that's a very complicated web to unweave. It is much easier to bring data together and to compile it than it

is to then try and disentangle it and figure out what goes where and where it came from. And so, again, speaking to government officials who are watching this happen, they're like, look, right now, these Palantir contracts are separate. They're happening in separate agencies. But it'd be really, really easy for the government as a client, the Trump administration as a client to merge it

And once that happens, like we don't know how you undo that at some future point. Or let's say some court rules against it in a year's time. Like we don't know how you ever undo that sort of thing. You also wrote about a product called Foundry. Can you describe it?

So Foundry is one of the main products that's offered by Palantir to organize data. It is, and I sound very Silicon Valley right now, but it is AI-backed. It is an AI tool. So it's intelligent about how it processes and organizes data. And one of the people I spoke to, he's a defense official in the U.S. government, described it to me like this. He said, look,

I'm, you know, I am a Luddite. I don't really understand data. I don't understand systems. But what I know is I can type something into it and it will present it in a way which is visual and easy to understand. It is made for people that don't necessarily know how to use technical tools to make it really easy to understand the data that they have in their own systems. Because I think one of the questions then is,

If you have a product like that, right, and it is going to multiple agencies with people who might not understand the reach of what they can do, like, does that create concerns, right? That you might have people within an agency using a product that is so higher tech than they are, for lack of a better way of putting it?

I mean, until now, that hasn't come up only because there's been such a need in government to kind of organize and understand their data. And I'll note again, until now, it's been really siloed. So if it got used at the CDC to help them organize the distribution of vaccines, there was this sense of like, well, that's great. We really struggled with this. We couldn't figure out which

areas needed more vaccine, less vaccine. We wanted to get them to areas that were COVID hotspots before we got them to areas where maybe, you know, it was really, really useful to organize and not just to, I mean, something else I found out, which was so interesting, was it does this kind of real-time analysis. So like,

if traffic was bad, if the road systems weren't working, if flights were delayed, it would be like, right, don't send the vaccines there because they're not going to arrive in time. Send them this other way. So you had all these people at the CDC specifically being like, wow, this is so great. This would have been so much harder for us to distribute these vaccines without this kind of data system in place. And that was, you know, that was the thinking in government for so long. Like what a great tool for us to use in our specific government agency.

What has been fascinating about reading your reporting and also some of the reaction to it is

is some concern from Palantir employees themselves about some of these projects. Thirteen former employees signed a letter saying that Palantir should stop working with the Trump administration. Can you tell me a little bit about this backlash within the company? I mean, I think that it's really interesting. So many people I spoke to joined Palantir because they believed in a certain part of its mission.

Maybe they believed in their anti-terrorism work or the CDC work or whatever, any number of programs they thought were worthwhile. And the Trump administration and the work that Palantir has done with the Trump administration has really shifted their thinking on this.

And I spoke to people who said things like, you know, we're going to become the face of the immigration agenda. We're going to become the face of families being deported from the United States. Because even though we are just a technology company, if our technology tool makes it possible for the Trump administration to find and deport people, that's going to be what people remember us for, right? That's going to be

ultimately our legacy. And there was a lot of, I mean, you have current employees who are posting, you know, within Palantir messaging groups saying like,

Like, what if they start using us, you know, okay, maybe you're anti-immigrant, right? But like, what if they start arresting people who are, you know, political dissidents? What if they start going after anybody that speaks out against the Trump administration? What if they start hunting down journalists? I mean, you've seen these, I've seen these internal messaging chats among Palantir employees where they're saying, we're scared about what this tool can be used to accomplish in the hands of

of someone who really wants to seek out any political rivals, for instance.

This is a little reminiscent of some of the things that happened around Google's contracts with the Pentagon in 2018 that led to some of the Google walkouts. Ironically, Palantir, I believe, also contributed to this, one of the data fusion projects that we're talking about. But this tension between tech employees and when and how that technology is used in the hands of the government.

Totally. And I think that, you know, you hear all these tech employees being like, well, we know we created these powerful tools, but we didn't think that someone would be in office that would use them in this way or that way. And I'm sitting out here in Silicon Valley, and it's actually interesting because now you have Google, Meta, Opening High, all these tech companies who I think under previous administrations had been really careful to say that they wouldn't allow their technology to be militarized.

are now changing their terms of contract and saying they will allow their technology to be militarized. So you are seeing something of a shift here of tech employees who are much more willing, I would say, for the most part, to work for companies that do work with the government, that do work with the military. At the same time that you're seeing Palantir, this tech company that has much, much deeper ties to the military and government really than almost anyone else out here in Silicon Valley,

questioning their own, I would say, legacy in all of this. I am interested in reading your work about how some of the data was used potentially by Doge employees. There's like grumbling that, oh, their security protocols weren't up to snuff. Yeah, I mean, that's a really important point that was raised by almost every Apollinaire employee I spoke to, which is that they've seen firsthand that

Some members of Doge have really kind of sloppy habits. They talked about things like bringing in cell phones or personal cell phones and downloading software onto government systems without going through the rigorous process that previously existed. And they're genuinely worried that...

Not only is this, let's say this goes forward again, because it hasn't happened yet, but let's say it goes forward and Doge, the Trump administration, compiles all this data of Americans in one place. That's like putting a massive bullseye, right? For any Russian or Iranian or Chinese hacker that's sitting there, you have now one. It's very valuable. Oh, yeah. You've literally taken all these different things and been like, guess what? You don't have to hack 20 different government systems. You just need to get into one.

And at the same time, we have these people in government who are doing things a different way and who are maybe not using the security protocols that have been put in place by previous administrations. And the worry of these Palantir employees is like, well, you know, who's going to be blamed when Palantir

When and if the hackers get in, it's not going to be the random Doge employee who brought in his personal cell phone and downloaded something onto it. It's going to be Palantir. People are going to say that Palantir was hacked, that Russians hacked Palantir, that Iran got into Palantir. And so, again, they're talking about like this reputational damage that they see is potentially happening to the company.

In response to Shira's reporting, Palantir published this statement on its ex-account. Quote, The recently published article by The New York Times is blatantly untrue. Palantir never collects data to unlawfully surveil Americans, and our foundry platform employs granular security protections. If the facts were on its side, The New York Times would not have needed to twist the truth. Unquote.

One of the really interesting wrinkles to me is watching how your reporting has rippled out. There are people like Nick Fuentes, the white nationalist sort of MAGA figure who posted his opposition to this idea. There's a little bit of a MAGA backlash. What do you make of that?

I mean, I think it's so interesting and it kind of reflects the history of this. Like I would say Democratic administrations, Democratic Party administrations have talked about compiling data in the past. They've talked, you know, we've heard of people in the Obama administration who are like, God, wouldn't it be so much easier for us to create some of these welfare systems if we compile data? And even back then you had Republican administrations.

you know, active party activists and Democratic Party activists saying, no, no, no, no, no. So many reasons this is a bad idea. So many reasons why this is scary and gives off Big Brother vibes. And once again, you're kind of seeing that happen where a lot of people are responding to this and saying, this is scary for us. You cannot put the genie back in the bottle once this happens.

I think Fuentes and some other sort of far-right figures in the United States have talked about, well, if Trump is one day not in office and the Democratic administration in office, how are they going to use the data? Some of the framing I've seen online since the story got published. So, yeah, I think a lot of people are very, very worried about centralizing this data because we live in a modern world where data is power.

In stories like this that talk about big data and often the intersection of data and the government, one of the questions that comes up a lot, and I get it from listeners, is, well, if I'm doing everything right, why do I care, right? Why do I care if they know about my student loans and who my cousins are and, you know, that my brother-in-law struggled with PTSD after military service and

And so I guess the question is, yeah, if you feel okay with the government having your information, should you worry? Why should you worry?

A lot of people probably don't have to worry, or at least they will initially not have to worry. But I would point them to the... And again, I'm not trying to say, I'm not trying to liken this administration to Russia or China or some other countries that I've worked in and I've seen firsthand how data gets used in those countries. But there are many countries in the world right now where the government will use data to coerce citizens into doing what they want.

in Russia, there are many accounts, many stories of a person who's never really done anything wrong in their life. Maybe they just had a student loan that they took out and they're still paying back or they have a sister-in-law with a health injury from when she was in the military and

that information, those pieces of information get used against them because somebody in government wants to buy the factory that they own or they want to force that person's son to join a specific unit within the army or whatnot. It often is hard for the human mind to think ahead to all the ways that things can possibly go wrong unless you've gone to one of these regimes where data is militarized and weaponized against its own population.

As you've said, this is something that you're reporting on. This is not something that is a finished product. I guess one question I have is how long might it take to build something like this? And how would American taxpayers know if it's finished?

So one of the reasons we wanted to do this reporting now is because, well, I actually had all these people from government reaching out to me and saying, we're really worried. We're seeing these systems being introduced and we're seeing these conversations happen, which indicate that this

This technology, Palantir's Foundry technology, is really rapidly expanding. And that's really like 90% of it is establishing Foundry in all these different government agencies. Because once you have all the data in these different government agencies, and by this I mean data on American citizens...

It's then really just about the Trump administration essentially signing a new contract with Palantir and being like, okay, this new contract is for now collating all those systems or like bridging all those systems together. And that's a fairly quick process. The long process, the hard process is putting Foundry into all the different government agencies. And

The analogy that was drawn for me by this one government official was like, look, it's almost like you're doing a group project and one person wrote it in Microsoft Word and one person used Google Docs and one person used WordPerfect and everybody used a different thing to write their essay on. The hard thing is getting everybody on the same system. Everybody's got it. But like, let's say you managed to convince everyone in your group project to write that essay in Google Docs.

All it takes is like a control C to copy paste that into the same place. That's that much easier to organize. You get no weirdness in the system. You've eliminated those glitches. And so I've now been thinking about it that way as like, you know, this technology is very rapidly going into these different agencies and

As taxpayers, as Americans, we're only going to know that it's collated when I think really when the Trump administration comes out and says like, hey, we're signing a new contract. We're buying this new, you know, we're establishing this new thing with Palantir and all this will be now collated. Sure, Frankel. Thank you so much for your reporting and for talking with me about it. Thank you so much for having me on.

Shira Frankel is a tech reporter for The New York Times. And that is it for our show today. What Next TBD is produced by Patrick Fort and Shana Roth. Our show is edited by Evan Campbell. TBD is part of the larger What Next family. And if you're looking for even more Slate podcasts to listen to, check out Thursday's episode of What Next. It's about a torture victim who was planning to seek protection in the U.S., only to have her right to do so quietly erased.

All right, we will be back on Sunday with an episode about one of the most important people in Washington who you've probably never heard of. I'm Lizzie O'Leary. Thanks so much for listening.