Why Palantir’s data domination plot isn’t working out
31m 54s
The podcast discusses Palantir, a data analysis firm co-founded by Peter Thiel, and its controversial integration into the UK government, particularly the NHS. The company, named after Tolkien’s seeing-stones, has been described as a "corporate bogeyman" due to its ties to US military and immigration enforcement, as well as its opaque technology. In the UK, Palantir secured a £330 million NHS contract for its Federated Data Platform (FDP), which aims to centralize health data. However, investigative journalist Peter Gagan reveals that the software is often impractical, with NHS analysts complaining about slow performance, inability to edit proprietary code, and poor support. Gagan also highlights Palantir’s lobbying tactics, including hiring Peter Mandelson’s firm, Global Counsel, to gain undeclared access to ministers and officials. The company’s "land and expand" strategy—starting with small contracts like a £1 COVID data deal—has led to significant vendor lock-in, making it difficult for the UK to disengage. Critics argue that Palantir’s technology, which supports government surveillance and military operations, poses risks to democratic control and public trust. While other countries are retreating from Palantir, the UK appears increasingly dependent, raising questions about transparency, accountability, and the influence of private interests on public institutions. The episode underscores the need for greater scrutiny of such partnerships and the lobbying that enables them.
[Music] Hello and welcome back to the bunker, your daily required intake of news with other nonsense, I'm Andrew Harrison. Today, the data giant Palantir, it's been described as the scariest company on the planet, is our government too close to it and what is it doing with the NHS? The data analysis company co-founded by infamous libertarian and democracy skeptic Peter Teal has become a corporate bogeyman for liberals, progressives and people who listen to podcasts like this and also to many MPs who this month urged the government to cancel Palantir's 330 million pound contract with the NHS, citing low public confidence in the company. Palantir is famously named after a magic long distance communication crystal from Tolkien, it means watch over from afar, all these guys have Tolkien for some reason and it has been accused of infiltrating the British state. The science, innovation and technology committee has warned, the Britain's increasing reliance on Palantir represents an unacceptable point of weakness, essentially. We could become dependent on a company whose technology is obscure and hard to disconnect from and it's also deeply integrated into Donald Trump's immigration crackdown and the US military, it's been condemned for expanding government surveillance through AI and facial recognition. So what is the British government doing dealing with these dangerous libertarians and with technology whose capabilities both outside of democratic control because it's at transnational and far ahead of the laws ability to understand it let alone regulate it. Journalist Peter Gagan, author of Democracy for Sale, Dark Money and Dirty Politics, investigated Palantir for an in-depth feature in the London Review of Box which is out now, it's absolutely enthralling and will link to it in the show notes. He's here with me today, hello Peter, how you doing? Hello, thank you so much for having me on the show. Not at all, well listen, first up, I mean for a lot of people while I'm sketched out what Palantir is, you know far better than I. What exactly is this thing because before we started recording you said it presents as a technology company which is actually something quite else. I think it's a really good question and I think in some ways you have to spool back to where Palantir started. So Palantir started basically 20 years ago. It was set up above a sushi restaurant in Palo Alto, set up by Pichotille as you mentioned, famously he'd been busy in La Mosque, set up PayPal, you know, libertarian democracy skeptic and Alex Carr, his former roommate from Stanford who for a long time was kind of presented as the democrat who describes himself as a socialist. His views of changing his fairytale might get into that. But basically they came out of the war on terror, they were basically a kind of a software company that said what we can do is we can take all this data, you know, so much data, you know, and listen to America, American government dollars data, all these different places, we can bring it all together, we can harness it. And in the context in which it was started, a lot of it was about we can harness it for military, we can harness it to try and you know, to surveil people, find out military targets, etc. But also for commercial use too. So that's where it palent here emerges out of that world. And what it's selling and frankly what it's still selling is this supposedly amazing ability to take data from different places, bring it all together and allow you to do whatever you want with us. It's kind of the fantasy of total information awareness, isn't it? If we knew where everything was and how it all related, we'd be able to see what everybody's doing, but also predict things. And also I guess, you know, kind of without wanting to see a nickel about it, you supposedly could find it efficiency, supposedly you could find new things to do with the stuff that you've got. What is it actually offering now though, 20 years down the line? As I said, very deep relations with the British government and other governments around the world, which you can get into and over when Palentay kind of opens a suitcase now at the presentation. What is it selling? Well, this is it and it's selling a lot more than what it used to sell. So it used to be selling effectively, you know, military, military grade intelligence, ability to bring it all together. It's now in lots of different things. We'll talk about the NHS more, I'm sure, one of the big things is health data, but it does lots of other stuff too. It works for police, the stuff like predictive policing. It has got so many contracts, a good contract with financial conduct authority with the atomic weapons inspector in the UK. It's effectively a massive and this is when the great ironies of the company, I think because of Teal's background, Teal is a libertarian who's very anti-destate. It is a massive public procurement machine. What it's trying to do is who've rubbed government contracts and it's so far working very well because it's market cap, last time I looked at about 325 billion dollars, which is about 150 times its earnings. The economist has called it the most overvalued company in the history of kind of contemporary economics. But what do you get? Well, what they're telling is still the same play effectively. We can take data and we can do things where other people can't do. Right. You dug deep into this and they've included everything from attempting to doorstep balance here at their offices, which are inevitably kind of code-named after something from Tolkien. When are these guys going to go up? I can't believe I've said that, but as you're digging into it, what's the surprise you? Because it's weird and often first blush. I think what's surprising is that there's a couple of questions that started with me and my colleague Lucas, I mean at Democracy for Sale. We've been looking at Pounder for a long time and I felt like there was a lot of talk about Pounder and for a long time I kind of was a bit like I read the biography of Peter Teal that Max Chey from brought out a few years ago, but also as regards Pounder itself, I felt a bit like I don't really know you hear all these things and they say so much. I was a bit like I don't really know what I think about this. And I guess the two questions we were trying to answer was one, how did they get so am I actually in the British state? How did they get so involved? And the second question was like frankly, does this work? Does the case from the you know, and this was kind of motivation particularly by conversations Lucas was having with NHS whistleblowers and data analysts who were just complaining about Pounder a lot. And effectively what the big thing for Pounder in the UK, the big flagship thing it has is what's called the federation data platform, the FDP. If you hear about NHS data, Pounder, what you're talking about is the FDP. And one of the things I enjoyed about writing this piece actually is actually understanding what that FDP is. I'd heard it written about so often and getting sources to basically take photos of what it looks like and effectively to dashboard. It's a series of dashboards that tell you about different things that are going on. You know, it'll tell you about what's the occupancy ratio of beds in a hospital, tell you what appointments are coming up. It's supposed to basically exist as a space in which all of this data that's held in disparate parents of the NHS, whether it's in trusts or hospitals or doctors or pharmacists, can all be brought together in one place and utilise it better. I think everyone could agree that sounds like a good idea. It's a good idea to get all of our data together. But what was really interesting is the more I started talking to people within the health service and the more I started to poke with people who were using the software all the time, the more you realise just frankly the limits of what it can do from what we're trying to do. And it's a really good conversation with somebody who worked very senior in NHS data who explained to me, "Pounder is great if you're making widgets. If you're trying to produce widgets, their software is very useful because it's a uniform thing that will come out the other side." The problem is day shift to do with healthcare is actually quite often it's quite qualitative. It's also very messy. It's in lots of different spaces, lots of different things. And effectively trying to pull that into one centralised software because the problem in some ways is that what we have is Palantir's products. We bought their product. It's called Foundry. We use it. It all gets pulled into Foundry. And that actually doesn't really work for a lot of things. It takes a really long time to load because it's proprietary software, NHS analysts can't edit it themselves. They have to wait. There's a really interesting bit in the piece where NHS analysts have to ask Palantir if there's a problem. So they need to fix it. And at one point an analyst puts in as a technician to help with something. And a forum is an online forum for this. There's no response after a week. So he posted a gif of a tumbleweed. So one of his mates starts coming. So there's still no response. There's pulling a gif of Mr. Bean. And you're not one of the dog looking out the car window, the background. And so it kind of gives you this sense. And that was I found really interesting. Like peeling behind the layers of all this kind of incredible eschatological kind of doom about the future of the world to ask, well how does this software work in practice? It's almost like they're selling a fantasy idea that's a little, almost going to be out of reach. The idea that if you can see everything, you don't understand everything. It's like, I don't remember that. The bit in Watchmen where he's standing in front of millions of television, Adrian, in fact, and now I've got full-in for what I can see what's happening in the world. But it just doesn't work like that. You can't fit that stuff in your head. And a dashboard doesn't fix it either. The thing that stands out in the feature is people saying it doesn't bloody work. They make us use it. It doesn't bloody work. And then we get taken into a NHS trust get together. And the first thing we're told is don't complain about it. It's really good. I don't want to hear any moaning. It's like every little bit of crap software you've been subjected to in your life. It does. And except for that it was added kind of video of scaryness. So in that conversation you mentioned, all of those people who went in there because that meeting for the NHS, you know, leaders to talk about it. It was in the early once the FDP. It took place in Panantir's headquarters in London and so forth. And it was going to be discovered right in the piece. It just signed an NDA to get in. So most times you don't actually have to do that. If you're going to get a braided at least you don't have to sign a way over saying I've never got any. I didn't make you son an NDA to come into podmasters. Maybe I should have done actually better control this conversation. And what's also interesting about it is the answer to problems becomes more data. And this is this is so one of the things is so because of the opposition back in 2023 to Panantir, been given the NHS FDP contract. There was a car ride for GP data. So GP data is an included. So you'll often hear people say if we only had GP data in there, it would be great. And Louis Mosley actually who's the head of Panantir UK, we need to give evidence that COVID inquiry, he talked basically about how Panantir could provide the operation system for the whole stage. So really what we need is more and more information that would make this work. Yeah.
But that actually was something that really stood out to me as well. The idea that the more of this stuff you buy, the more it becomes woven into the state. Is the kind of dream of Peter Thierry thinks that freedom and democracy are not compatible and is part of that community of people who thinks that soft corpse, sovereign corporations are going to replace the state? I think basically trying to not sell a tool for the state, but like a replacement operating system for the state. Not getting too conspiratorial about it is strikingly if you read the sovereign individual or something like that, William Rees, Mark's book. This is the kind of world that Peter Thierry describes. And effectively, there is an element of that. They talk about an operating system for the state. They talk about providing initially an operating system for the NHS. This is the language that is used. Like it is this idea and it's not a coincidence. Peter Thierry is not someone who likes the idea of the NHS. Like he does not like the idea of free of the point he is healthcare. Like he's someone who would describe that a socialised medicine if not much worse I'm sure he has. So these are not people who are by nature, particularly friendly towards this. And it's also striking, I think that Britain is now looking like an outlier. A lot of other countries are really stepping back from Palantir. The French, the Danes, many others have kind of gone, "Actually, do you know what? Palantir is now in the last six months, actually, no, since Trump got back in. Become really an arm of the Trump White House. It is explicitly saying part of its job is to protect American power and support American power. We're in a situation like we are now where it's really hard. I think the piece does describe why it's so hard for us to leave Palantir. Now it's in the NHS because we're now basically locked into it. It's what's called vendor lock and it's hard to leave. That's a real issue, I think. Yeah. I mean, I guess that raises two questions. One, if Peter Thierry is so skeptical to us at the NHS, why does he want to be in it so badly? I guess that question is rather easily answered. The other one is, why do we want to be in? Why are politicians so keen? To work with this person who doesn't believe in democracy, very probably doesn't think that socialised medicine is a good idea indeed, socialised anything. Why are they so keen to pile in and knit this into our state? I think there's two answers to the wider so keen. I think why is he wanting to do it? I think there's a pretty obvious one which is money and the kind of, the kicker for that is the NHS has got a particularly strong data set. The data to the NHS really is almost unique because of, in America, the closest you have is veterans data and that's kind of, it's very useful but it's got limits because they're all veterans. The NHS has a particularly unique healthcare data set. But in terms of the second question, why, one of the things I really wanted to answer in that piece was how did it get so involved in the British state. The answer to that, unsurprisingly, if anyone has ever read Democracy for Sale, or my book, or newsletter, is the influence of lobbying. So, balancing our hired Peter Mandelson's lobbying firm Global Council back in 2018, explicitly to kind of reposition themselves because it's a pretty bad track record in the state. And they were seeing it very close to the CIA. The CIA was involved in setting them up. I think the shia is a front for Palantin, so you're quite a good rival at Bonn Mouche. But effectively, under Global Council, Palantin, we're able to really get a huge amount of influence and a huge amount of access and all of it really underclared because it didn't have to declare them as a client. And so, with Mandelson, and it's interesting talking, the Global Council doesn't exist anymore. It went bankrupt earlier this year, it went to Boston after the revelations in the Epstein files. And it's really interesting talking to former Global Council staff about how interested Peter Mandelson was in Palantin as a contract, how interesting it was in them as a client. And back, if people remember last year, when Kierstarmher feels like a million years ago, when Kierstarmher went to visit the White House, with Peter Mandelson, after that, they did go to Palantin's showroom in DC. He met with Alex Carr, he met with Louis Mouche, they starmered. The British government refused to say what they talked about, but afterwards, Palantin here's had even more UK government contracts. So I think that's a really important part of the story. There's two bits that we're amazed by. One is that, wasn't Mandelson acting for Global Council while the ambassador? So he's working for us and for the lobbyists who are trying to act as intermediaries between us and this company. And the other is the role played by not just Global Council but other lobbyists, primarily just to the longer the contact, right? So Palantin turned to it number 10, it's got to be declared. Here's Alex Carr, Peter Dale, probably not him. He lives in a hyperbaric chamber on the moon or something, but like it would have to go into minutes. But when it's just a conversation over dinner after the event with a lobbyist, then that doesn't get minted and you don't know what to talk about. Yes, this is exactly the big major issue. I think the story really shows just how Palantin used that. So what Global Council would do is they do stuff like they put on an event, say it was about say, for example, AI and healthcare and you bring along a minister, you bring along senior civil servants, you bring along policymakers, you'd have Palantin and people in the room, you'd have the head of Palantin, they all get to meet you, should they all get to talk, they get to swap numbers. All of that currently doesn't get classed as lobbying under the UK, under the UK lobbying act. So, interestingly, last week, this thing called the Ethics and Integrity Commission at Labour, it's set up because ironically, Mandelson's fall has precipitated finally some conversations around us because exactly that what had was Peter Mandelson wearing. He owned at least a fifth of Global Council, why he was the ambassador, these are his clients in whose interests is he working for. So they have recommended actually that all of this should be declared. And I think it's going to be very interesting if Andy Burnham actually picks up those recommendations because for years we've had this, this is how these lobbying firms work and Global Council are not the only ones. They are able to structure their interactions between clients and government and ministers in a way that doesn't hit these regulations. So you and I would think that like, if you're going to meet a minister and you're working for these clients, well surely that's got to break the lobbying act. No, you have to be talking to a minister, senior civil servant, about a specific client and an issue of interest. So just even just generally saying how great they are, it doesn't really cost us. Like having a priest to intercede between you and the Lord really, that kind of thing. It's alright if you could do it that way. Well, I was reading this and it is an amazing piece. I started to think, is this actually quite a mediocre tech company that's also a world-class lobbying outfit? That's what they're really good at because you describe this fantastic strategy called "Land and Expand" where case in point, it's COVID, Boris Johnson and Dominic Cummings have an off-the-books meeting with Taylor in Downing Street, not in his type of a brick chamber. Palantay had charged £1 to get the contact of managed NHS data and they end up with a 23.5 million pound contract which grows and grows and grows. So they're very good to get in the foot of the door for £ or less and then just colonising I guess. This is it, that's exactly what I landed in Expand. It's basically you give someone a taste and again you give someone a taste and they get used to your product. It just sounds a bit, I sell a little bit like I'm in the wire now. It's the nicotine model of technology, yes. So you get them into it and that's exactly it. They start using it and it becomes much harder from the get off because of this vendor lock and they're now using the product. What they're doing is on your software, it's much harder to leave. What's interesting and we write about it in the pieces, a much less talk to Bosch example this with Palantay was back in 2022 after the invasion of Ukraine, the then ministry of leveling up wanted to set up a system to try and kind of all these Ukrainian refugees were coming system to try and coordinate. So Palantay came in, offered for it didn't even charge a pound this time completely for free, decided to build a system. Ten million pounds were to contract later. The national audit office, people are going this, this doesn't really fit for purpose, it's not working. The national audit office looked into it, found it wasn't value for money, found the systems confusing to use. The ministry, the housing has since ditched the software as using its own in house software and area this year was bolsting of saving millions. Are there competitors to it? Are there other people offering this stuff in the same way? Or is it the one and only vendor of ultimate problem solving universal big data solutions dot com? No, I think what's interesting about Palantay is, because I've kind of touched on the piece as well, like some people call on Palantay evil sales force. And I think it's a pretty good way of putting this. Like they've got this malevolence self image and I think this is intentional. It is intentional to present themselves as this soul scary. If you talk to people in government, let's say, look, we don't like their politics, but they're kind of evil geniuses. And the last thing, what's striking when you start like picking some other ways, actually the product is pretty, in some respects, some of it's good, some of it's mediocre. The box stand, but it's free. Yeah, it's not, it's not, I think a lot of like the stuff that they do for us, the stuff that they do for Israel Defense Force, they're working for people doing evil things, frankly. In the context of what are places, often what they're doing is just, it's bog standard. And there's something about their ability to depict themselves in these kind of terms as these great, like they don't have a foundation layer model, like anthropic. They don't build hardware, like even Elon Musk does. They were software company and we put a lot of these things at Palantay, actually, in France, they came back and said, we were software company. It's up to NHS Trust, how they use us, that's up to them. And so I think that's kind of a telling response. They do seem to be, their name seems to ring out so loudly, but effectively what they are doing is building software. I've never done any kind of more complicated procurement than trying to fix me mobile phone card, drag myself right. But I've got mates who sort of send me in that world and they're kind of recurring complaint is that the sale will be made to a politician who's easily dazzled and won't be there for forever, particularly health ministers. Or it will be sold to somebody who might be running an aspect of the prisoner state or might be in some, they might have a bit of a home office and they want a flashy, fast, visible solution. So they buy at the top of the market from the recognizable name, just as much as I, clueless.
will go and buy the name of a stereo that I recognise. And that maybe that's how the kind of thing is organised. It's not so much about craft and a good product. It's about being known. It's a huge part of what we've called from people. This is a recurring issue when it comes to NHS data. So we've spent in the past, we spent billions trying to do this in a similar way. For almost the exact same reason, it was a great call from someone on the record in the piece who worked under Tony Blair and was on data issues from the NHS. And he was complaining at the time. We basically just hired these flashy American tech companies. They dazzle ministers, ministers they will want this thing and the thing they build and the thing that comes back time and again reporting we did on this piece is that it just really doesn't. It's not fit for purpose. It's kind of one-fit size fits all. The people at the call face trying to use it are putting their hands going. This isn't the way to do it. And actually strikingly, a lot of the NHS trust because one of the things that comes out in the pieces, a lot of NHS trust just aren't using the F.P. They're just not using it very much. And ironically it's the trust who actually had the weakest data systems they're using it probably the most because actually didn't have much to start with. But a lot of the trust who are more advanced and are using data systems that they've built themselves are saying, look, this just isn't as good as what we've got. They've built their own systems all from an open source and they're working better for them. And meantime, one of the things I find quite striking is actually the UK Statistics Agency is now investigating NHS England over NHS England's use of statistics to promote Pallanteur's product. They're saying there are these in serious questions as a piece in the FT at the weekend which very much supports our reporting too, which is raising serious questions about NHS England's own stats that they're using to try and support it. So it's some cost to the point where you actually jack in your own stats to make it look better. I want to ask you a bit about defence trust that I then maybe sort of try and defend the idea of Palancere a little bit. I mean, they're into everything. I read yesterday that they've just got the contract to manage Britain's private firearm database. It's only about 9 million quid. There's only about a handful of private firearms in this country, but they're doing that as well. How deep are they into the British defence industry? They're very increasingly is the answer. So they're kind of properly landed in London around middle of the last decade. And at that stage, the only work they'd read on the UK was for a GC HQ. So they've always had a big defence bit. They hired, and this is a big thing as well. They hired a lot of ex-defense people. So they've hired a lot of people from the MOD. They did that early, and they've, to quote your Irish phrase, done it early and done it often. You know, they've brought a lot of people in from defence industry. And that's been very successful for them. So they've got a lot of big defence contracts. They've got a big 240 million pound contract last year. Not long after that meeting between Karp and Mosley and Kyristar Mar they signed a 750 million pound deal with the MOD for the next five years where we've now committed to buying three quarters of a billion pounds for their tech. Interestingly, the MOD has since actually had to acknowledge that basically they're locked into Palantir that if they were to stop using Palantir on somebody's contracts, they would have to basically start all over again. So in some ways, what's happening in the NHS has been replicated in the defence space as well? I mean, for a lot of people it's concerning in that, as you described, the policy is very deeply woven into the Trump administration as the American military. And by definition, it's a big data company, the bigger the data, the better. I'm in a situation where our own military capability and our own data, because of the transnational nature of this business, are possibly open to countries which might not have our best interests at heart. I kind of read that in the piece, there's a national security issue. I think there is. I think if you look, if we were in a context where the Trump administration has actively called out to leave our government on numerous occasions, they are not our friends. And it's striking that a lot of people who, Britain, I think now, at least the current government, see themselves closer to, I think, we're clearly closer to Europe in terms of our views on defence, on basically the world and liberal democracy. Most of those states in Europe are really close and door on Palantir. The Danes, the Dutch, the French are all really, really wary of Palantir and getting more wary, and are all committing various things to kind of not take on contracts. I think Britain is now starting to look like it's in quite a vulnerable position, because data is as you're right. Data is what Palantir thrive on. And there is still questions. Questions have been raised about access to data from Palantir staff to health data, et cetera. And given the nature of Palantir as a business, I think it's not unreasonable to be concerned about it. And also, there's the kind of symbiosis between large tech companies and the Trump administration in that Trump frankly acts as their enforcer on the world market, tariffing countries that attempt to tax them. We want to tax American tech companies what we say is fairly in this country, but we deem it because we fear the tariff, and that's a new thing in the world. That's not something to happen before. And we're a relatively large country, a relatively wealthy country. What if you're not? What if you fear the tariff even more? Yeah, I think there's a lot of that too. And you can see, the politics of Palantir is really fascinating. And the FT, there's an element in which is Palantir overreaching. There's a very good piece in the FT maybe last week in the financial times where they kind of looked at Palantir and the US. And basically asked that question, is Palantir's male evidence self image effectively? Well, I write one in this piece. Is that doing a company of damage? And fold in the US, but also in the UK, people are leaving Palantir. Palantir, people in Silicon Valley say you can't get a date for your work for Palantir. And Silicon Valley is not exactly full of virtuous angels, but even within that context. And they've set themselves up so much as the enforcers of Trump's agenda, whether it's I, they're very, very close to Israeli government. They took out a full page ad, we stand with Israel after the Hamas attacks in 2023. They've held a board meeting in Tel Aviv, like again, as a kind of show solidarity, but they are pushing themselves into very political positions. And ironically, having for a long time sat between the Democrats and the Republicans, Teal, been in Republican, Teal is very close to JD Vance. I didn't put that in the piece. He was the first big backer JD Vance car, gave money to Democrats, they sat between them. Now, Carp is giving money to Trump. He gave Trump money to Trump's inauguration. Teal is backing Republicans in the midterms. And Democrats on mass are saying we want to be anything to do with you. And if the Democrats get back into power or a Palantir stock price, I wouldn't be surprised if it starts to go in the other direction. Now, I have to devil's advocate this, right? A lot of people would say some of the finest people, very, very good people, great people. Some people would say that, like, this is going to happen anyway. We are moving into a data world, if I were already there. And if we want to maximize the opportunities that it affords, if we want to run our government efficiently and innovatively, if we want to have an NHS that is going to be able to more successfully predict and prevent diseases, we have to embrace this. And they would probably also say that government itself is a really bad record on generating its own tech. It has an equally bad record of pecure and get from outsiders. But frankly, are you likely to get your solution from a government-backed government control entity or from a market-produced entity? They would probably argue that the latter is the case. Also, what they're saying is, this is coming anyway, we're going to have to do it. If Palantir, the only game in town, are there other ways of doing this? I'm not a data scientist, but from talking to experts around, as I was really struck, there is other ways of doing this. I think that is the one thing that does come across. There's other providers who probably have more experience. One being Larry Allison's, Aureka, Larry Allison's obviously very close to Trump, too. I've written about him as well in connection with 20 Blowns, too. But frankly, Aureka can do a lot of what Palantir does, the similar ideas. There's other off-the-shelf aspects like that. There is open source options as well. What's striking is that the tender process, I'll write by this in a bit of depth in the piece, for the FDP for the federated data platform. The process of love being around, I think it's fair to say, did produce a tender that was very friendly towards Palantir. So what we wanted to do with our data seemed to be very much reverse-engineered from what Palantir does and can do. And I think that in some ways was a fundamental flaw in a lot of this. Because we started almost from the premise of, what can these guys do for us? And that becomes the thing that we look to do, rather than what do we actually need. I think we all agree. And can all agree that we need to harness a data better. I think there's justifiable concerns about the type of corporates who we are entrusting someone out with. And the safeguards around them, especially in the current climate that we're in. Just to wrap up, then, we're probably going to have a new private minister. By the time people hear this podcast, baby, I mean, certainly within a week. Are you expecting a change of direction from Andy Burnham's government? Because he's trying to put clear difference clearly what he's called it between the Star Media Administration and himself. Are you expecting something different? I think it's a really interesting question. Now, two select committees in the last month of come out and said we should exercise what's called the break clause, which had effectively end the contract with Palantir in 2027. There's a lot of disquiet in parliament around Palantir. There's a lot of disquiet amongst people too. So there's a lot going on there. There's the kind of evidence that this piece shows about questions about how the FDP is working, both for the NHS and how trust are using it, and is it working for us? I thought it was really striking, but a week ago, there was a piece in the telegraph. And Palantir has definitely become a bit of a culture war issue, I think, in the UK. The right wing press, the telegraph in the time to particularly become very gung-hole, very pro-Palantir in a way that feels, it's become quite cultural-warified. And that's not very time-sy, is it? Because the times usually hate tech, because Rubai hates tech. But this is really, if you get Google Palantir in the time, it's just like every couple of weeks, there's another story about what Palantir could do for us. So that's really striking. And I was struck that it was a piece in telegraph about a week ago that said, "Allys of Burnham say he's going to ditch Palantir." And there wasn't many quotes in it. Allys of Burnham has become a phrase that I feel has kind of lost a lot of meaning. So it'll be interesting to see if he does a Prime Minister. But his box would be very quickly said that wasn't true and I think
I feel like it's going to be, I think this is a big test. I don't know if they've made any decisions on this yet. I think who he appointed to health actually will give us a big signal on that. I think why streaming was very sympathetic to American Silicon Valley solutions depends who I think who comes in next. But I think it's something that he will come under pressure about especially because there's a body of support in parliament that isn't just in the Labour Party, it's raised concerns about it. - He's just going to find one in Manchester, isn't he? It's the same, but it's from Manchester, therefore, it's better. - Well, it's interesting, sorry, we go to second, but I saw it in the piece. When a chief data scientist from Greater Manchester, trust actually talks about how he gave evidence to the select committee last month and says, we don't use parliamentar because what we have works so much better. So there might be someone there he could talk to about us. - My god, my stupid joke comes true. Peter Gagan, thanks so much for joining us, been really fascinating. - Thank you so much for having me, it's been an absolute pleasure. - And we've only just scratched the surface of this absolutely geodroping feature. The full thing is on the London Review of Box website and we'll link to it in the show notes. And while you're there, don't forget to click on the link to Patreon, the crowdfunding platform that enables you to support the projects you like. Patreon is the reason that we're here and the reason we can pay our people fairly, build studios, all that stuff. If you're a long-term bunker listener or brand new, please do think about back in 'cause it really helps. You'll get every addition, ad free and early, plus gorgeous merchandising and the deep satisfaction that comes from defending independent podcasting. Thanks for listening, we'll see you next time. (upbeat music) The bunker was written and presented by Andrew Harrison and produced by James Liddell, with audio production by Robin Leban, art by Jim Parrot and music by Kenny Dickens. The managing editor is Jacob Jarvis, group editor Andrew Harrison, and the bunker is a podmaster's production. (upbeat music)
Podcast Summary
Key Points:
Palantir, co-founded by Peter Thiel, is a controversial data analysis company with deep ties to government and military operations, now holding a £330 million NHS contract in the UK.
The company’s flagship NHS product, the Federated Data Platform (FDP), is criticized by analysts as inefficient, proprietary, and difficult to use, with limited practical benefits for healthcare.
Palantir’s growth in the UK is linked to lobbying, particularly via Peter Mandelson’s firm, Global Counsel, which facilitated undeclared access to ministers and officials.
The company uses a "land and expand" strategy, starting with small contracts (e.g., a £1 COVID data deal) and expanding into larger, long-term agreements, creating vendor lock-in.
Critics argue Palantir’s technology, tied to Trump-era surveillance and military applications, poses risks to democratic oversight, while other countries are stepping back from partnerships.
The UK government’s reliance on Palantir represents a potential point of weakness, with concerns about transparency, accountability, and the company’s libertarian ideology.
Summary:
The podcast discusses Palantir, a data analysis firm co-founded by Peter Thiel, and its controversial integration into the UK government, particularly the NHS. The company, named after Tolkien’s seeing-stones, has been described as a "corporate bogeyman" due to its ties to US military and immigration enforcement, as well as its opaque technology. In the UK, Palantir secured a £330 million NHS contract for its Federated Data Platform (FDP), which aims to centralize health data.
However, investigative journalist Peter Gagan reveals that the software is often impractical, with NHS analysts complaining about slow performance, inability to edit proprietary code, and poor support. Gagan also highlights Palantir’s lobbying tactics, including hiring Peter Mandelson’s firm, Global Counsel, to gain undeclared access to ministers and officials. The company’s "land and expand" strategy—starting with small contracts like a £1 COVID data deal—has led to significant vendor lock-in, making it difficult for the UK to disengage.
Critics argue that Palantir’s technology, which supports government surveillance and military operations, poses risks to democratic control and public trust. While other countries are retreating from Palantir, the UK appears increasingly dependent, raising questions about transparency, accountability, and the influence of private interests on public institutions. The episode underscores the need for greater scrutiny of such partnerships and the lobbying that enables them.
FAQs
Palantir is a data analysis company co-founded by Peter Thiel and Alex Karp, started about 20 years ago above a sushi restaurant in Palo Alto. It emerged from the war on terror, offering software to consolidate and analyze vast amounts of data for military and commercial use.
Palantir is controversial due to its ties to Peter Thiel, a libertarian democracy skeptic, and its deep integration into UK government, including a £330 million NHS contract. Critics worry about dependency on a company with obscure technology, its role in US immigration crackdowns, and lack of democratic control.
The FDP is Palantir's flagship product for the NHS, designed to bring disparate health data from trusts, hospitals, and pharmacies into one centralized system via dashboards. It aims to improve efficiency but has faced criticism for being slow, proprietary, and difficult for NHS analysts to modify.
According to journalist Peter Geoghegan's investigation, many NHS analysts find the software limited and cumbersome. It takes long to load, issues require Palantir's intervention to fix, and it struggles with the qualitative, messy nature of healthcare data, leading to frustration among users.
Palantir used lobbying firms like Global Counsel, led by Peter Mandelson, to gain access to ministers and policymakers through events and informal meetings. This allowed them to build relationships and secure contracts without triggering lobbying regulations, a strategy known as 'land and expand'.
'Land and expand' involves offering a low-cost or free initial contract, like a £1 deal during COVID, to get a foothold in an organization. Once inside, they expand their presence and contracts, making it hard for clients to leave due to vendor lock-in.
Chat with AI
Loading...
Pro features
Go deeper with this episode
Unlock creator-grade tools that turn any transcript into show notes and subtitle files.