Collective Property, Private Control

You may remember​ Palantir as the company that was given access to all of NHS England’s data in November 2023, in order to create a Federated Data Platform. The cost was £330 million – the largest NHS technology contract to date. Palantir’s first sales pitch to a UK agency came much earlier, in 2008, when its representatives gave a demo to an enthusiastic audience at GCHQ. At some point after that, GCHQ appears to have acquired Palantir software, despite misgivings about the use of a commercial product to store national intelligence. Project Gotham, Palantir’s main data-mining application, is used by the US military and security services and by approved allies, including the UK and Israel.

Data is the most valuable commodity in Silicon Valley. Technology firms collect it on a vast scale; find patterns in it to make billions from advertising; use it to train artificial intelligence models; and sell tools for analysing it to anyone who will pay. Palantir claims that its software allows customers hampered by incompatible digital platforms – an old IBM mainframe running COBOL programs, for instance, alongside a network of computers with UNIX operating systems, Excel spreadsheets, SQL databases and so on – to find information they never knew they had. Its name comes from the indestructible crystal balls in The Lord of the Rings. The wizard Saruman’s misplaced faith in this uncertain technology, which provides only a partial view of events, eventually leads to his downfall.

The head of Palantir UK is Louis Mosley, a Tory activist and the grandson of Oswald Mosley. The company’s seed money was furnished by In-Q-Tel, which functions as the CIA’s private venture capital firm. One of Palantir’s founders, the billionaire Peter Thiel, described Christopher Columbus as ‘the first multiculturalist’, accused Aimé Césaire of not understanding the transcendental value of The Tempest and advocates for cyberspace, outer space and sea-steading as routes of escape from ‘the unthinking demos’. His co-founders include Joe Lonsdale, better known for footing the bill for the ‘anti-woke’ University of Austin (its first cohort matriculated last autumn), and Alex Karp, Thiel’s classmate at Stanford, now Palantir’s CEO and the author, with Nicholas Zamiska, of The Technological Republic, a critique of Silicon Valley’s unadventurousness that effectively serves as a sales brochure for his firm.

Karp prides himself in being a rebel and disruptor and describes himself as a ‘socialist’ (his socialism may need some sort of qualifier). He is, more than anything else, a consummate salesman. He appears on Palantir’s earnings calls dressed in standard-issue white T-shirts and with a quasi-punk haircut. He rattles off quarterly earnings and throws out jargon like ‘ontology’ and ‘user-centred machine learning’ as if Palantir invented them. (Object-oriented ontology was already in vogue in the 1990s with the spread of the C++ programming language; user-centred machine learning simply entails a human user tweaking an algorithm by designating elements of the data as relevant or irrelevant, correct or incorrect – a bit like choosing attributes for a potential partner on a dating app.) At the end of one earnings call, which you can watch on YouTube, Karp throws his arms above his head and announces:

We have dedicated our company to the service of the West and the United States of America. And we’re super-proud of the role we play, especially in places we can’t talk about … 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.

In the same call, Palantir’s chief technology officer, Shyam Sankar, argues that the US is at war with China, that China is using fentanyl as a weapon against US citizens, and that the Belt and Road Initiative is a means of forcing other nations into ‘indentured servitude’ to the Chinese Communist Party. Sankar regularly denounces the Pentagon’s ‘monopsony’ power, the ‘deep state’ and the dominance of the ‘primes’ – the five major aerospace and electronics contractors: Lockheed Martin, Boeing, Northrop Grumman, Raytheon and General Dynamics – which he characterises as lumbering giants holding back the US military’s technological capabilities.

This complaint is also central to Raj Shah and Christopher Kirchhoff’s Unit X. Shah is a former fighter pilot who, while bombing Iraq in 2006, discovered that US air force jets lacked widely available and inexpensive GPS mapping software. Kirchhoff, whose Cambridge doctoral thesis was called ‘Fixing the National Security State’, worked as an adviser to the chairman of the Joint Chiefs of Staff and later to the National Security Council under Obama. The two men met at the Pentagon in 2015 and the next year joined the Defence Innovation Unit set up in Silicon Valley by the then secretary of defence, Ash Carter. In his previous role as undersecretary of defence for acquisition, technology and logistics, Carter had chafed at the Pentagon’s complex, incongruous and impenetrable procurement practices.

Obama had attempted to woo West Coast technology entrepreneurs without much success. The former CEO of Google, Eric Schmidt, was the main exception: he endorsed Obama in 2008, managed the data analytics for his 2012 election campaign and was described as the administration’s chief technology officer. Just before leaving office in 2016, Obama appointed him to the Pentagon’s Defence Innovation Board. Carter also invited In-Q-Tel to use Pentagon money to acquire new technologies. In the mirror world of US defence tech, it seems that the Pentagon is wooing a reticent Silicon Valley, not the other way round.

The US Department of Defence is a leviathan. In 2024, its budget was $800 billion, nearly 40 per cent of the global spending on defence, and more than that of the next nine countries (including Russia, China and Saudi Arabia) combined. As if this wasn’t enough, Trump has increased next year’s budget request by 30 per cent, to more than $1 trillion: $962 billion for the Pentagon and $107 billion for the Department of Homeland Security. This doesn’t include the $100 billion allocated to eighteen federal agencies dedicated to intelligence gathering.

The structure of the Department of Defence can seem impenetrable. There are deputy secretaries, under-secretaries and assistant secretaries to whom various agencies and field activity offices report. The heads of the army, the navy and the air force are members of the Joint Chiefs of Staff on the military side. The Combatant Commands under the Joint Chiefs are largely organised around geographic areas, sitting alongside the Special Operations Command, Transportation Command and Strategic Command. The department employs almost three million civilian and military staff. This doesn’t include contractors in US or overseas installations, or those who provide products and services to the Pentagon. More than 170,000 active-duty troops are stationed at 750 US military bases overseas. The department is the largest institutional consumer of petroleum and producer of greenhouse gases in the world.

Since 2001, the Pentagon has spent $14 trillion, between a third and a half of which was spent on contractors. In 2023 alone, 55 per cent of defence appropriations were outsourced. A good chunk of the money was spent on LOGCAP, the Logistics Civil Augmentation Programme, set up under Reagan to privatise military food provision, construction, civilian base maintenance and transportation. The beneficiaries are firms such as Halliburton, once chaired by the former vice president Dick Cheney. Some of the payments are to companies that supply overseas ‘force protection’, as Blackwater did during the war on Iraq. In 2022 Pfizer received $16.7 billion to provide Covid vaccines for US forces. But the rest goes to the development of weapons and technologies of war – tanks, aircraft, aircraft carriers, software, electronic equipment, satellites etc – by the ‘primes’. Silicon Valley wants a piece of this bounty.

When​ colonists in North America began to mobilise against British rule, the army found itself short of the engineering skills required for fighting a war. Many French officers joined the Continental Army as sappers and miners, and as builders of fortifications, trenches, naval batteries, ammunition stores and protective walls. Even before the colonies declared independence, George Washington ensured that the Continental Army had a chief engineer; the founding of the US Army Corps of Engineers followed that of the United States itself by just three years.

In 1802, Congress authorised the founding of the US Military Academy at West Point and the Engineers’ Corps was tasked with supervising it. In the following century, the corps was at the heart of US colonial expansion westwards and into the Caribbean, as well as being central to wars against Britain, Mexico, the Confederacy and Spain. Army engineers built aqueducts, dams and canals, dredged and fortified rivers, built roads and ports and harbours, and eventually provided support to railway development across the continent. In the so-called Indian Wars of the 19th century, infrastructure was designed simultaneously for civilian use by settlers and to aid the enclosure, expropriation and pacification of Indigenous communities.

The manufacture of armaments was a different matter. The colonial reliance on foreign manufacturing and materiel was anathema to the military men who were procuring the weaponry for the new republic, even as they made use of French methods of military engineering. Federal arsenals and armouries at Springfield, Massachusetts and Harpers Ferry, Virginia had foundries for the manufacture and storage of arms. The Ordnance Department, mandated by Congress to oversee the armouries, was soon given authority to procure from private firms. The historian Merritt Roe Smith has shown that it oversaw the standardised and automated industrial processes that came to be known as the ‘American system of manufacturing’. Novel methods of production entailed the ‘division of labour and application of machinery in the production of firearms with interchangeable parts’. The federal armouries competed against private weapons manufacturers and supposedly acted as the standard against which the probity of suppliers’ estimates could be measured.

Shipbuilding was another area in which US military investment underwrote the expansion of heavy industry. By the end of the 19th century, the navy had phased out wooden-hulled ships in favour of iron-clad and later steel-hulled vessels. Initially, the construction of such ships in federal yards was possible only by procuring high-grade steel from Britain. But Congress wanted to facilitate local industry. Senator John Miller of California, a former general in the Union Army, argued in favour of ‘constructing American men-of-war from American material, by American workmen, to be manned by American seamen, and to be used in the service of the government of the United States’.

In response, the Department of the Navy, seeking to build up a fleet that would enable it to seize islands in the Caribbean and Pacific, cultivated relationships with Carnegie Brothers and the steel companies Bethlehem and Midvale. The historian Benjamin Cooling has described the four decades before the First World War as ‘the birth of the US military-industrial complex’, and traces its characteristics – ‘kickbacks, cost overruns, favoured contractors, political interest in defence-related industries’ – back to this early marriage between US navy procurement and firms engaged in steel production and shipbuilding.

Some of the oldest and best-known American corporations got their start by feeding the insatiable hunger of the US military. DuPont was founded in 1802 at the behest of Thomas Jefferson to manufacture ‘black powder’ (a mixture of saltpetre, sulphur and charcoal) for the army. Over the next hundred years it specialised in gunpowder and explosives, until diversifying in the early 20th century into the production of other chemical compounds and later into nylon and rayon textiles. When industry was mobilised during the Second World War, DuPont’s chemical engineers were put in charge of the production of the plutonium that was used on Nagasaki in August 1945. Dow Chemicals, founded in 1897, was also indebted to warmaking for its fabulous profits. It provided magnesium (used in the manufacture of plane hulls) and styrene (used to make synthetic rubber) to the Department of War, enabling massive corporate expansion. The air force used Dow’s herbicides and defoliants, including Agent Orange, to destroy jungle cover and food crops used by guerrillas in Vietnam.

Napalm was developed in 1942 at Harvard by a team of scientists from the US Chemical Warfare Service. It is a sticky gel containing a gasoline-based incendiary, which clings to and burns whatever it touches. CWS scientists tested the viability of napalm as an airborne weapon by rigging bats with tiny napalm bombs and letting them loose in various testing grounds in the US. Once it was considered deadly enough to be of use, napalm was manufactured by Dow in large quantities and then deployed in the firebombings of Berlin and Tokyo. Robert Neer, a historian of napalm, quotes a Japanese medical professor describing the carnage: ‘There was no one to rescue. If you touched one of the roasted bodies, the flesh would crumble in your hand. Humanity was reduced to its chemical properties, turned into carbon.’ Dow’s first subsidiary outside the US was established in Japan in 1952, where it supplied the US in its war on Korea. The US military also used napalm in Iraq in 1991 and during the so-called war on terror.

In the later 20th century, the relationship between the military and industry became closer still. By the 1950s, as Roe Smith shows, military enterprise was behind the development of ‘computers, sonar, radar, jet engines, swept-wing aircraft, insecticides, transistors, fire and weather-resistant clothing, antibacterial drugs, numerically controlled machine tools, high-speed integrated circuits [and] nuclear power’. The Pentagon not only funded the development of new products but also controlled the diffusion of knowledge about new technologies: the attendees at Bell Labs’ first industry-wide demonstration of the transistor in 1951 were vetted by the Department of Defence. Transistors were particularly important in the following decade because they were integrated into long-range missile systems.

Space and nuclear technology were the engines of the Cold War missile race between the US and Soviet Union. The Pentagon saw space as a new frontier for reconnaissance and surveillance satellites, and for experiments in meteorology and communication, supersonic flights and high-altitude aviation. US Brigadier General Homer Boushey (a name ripped from Dr Strangelove) gave a speech to the Washington Aero Club in 1958 about the merits of establishing ‘a retaliation base of unequalled advantage’ on the Moon, from where missiles, hidden on the dark side, would zero in on the Earth. The US was already the only country to have used nuclear weapons in war; that same year it secretly tested a nuclear weapon on the edge of outer space.

Around the same time, the Army Ballistic Missiles Agency, the Jet Propulsion Laboratory and other military bodies gave a briefing about their plans for space colonisation to Boeing, Republic Aviation, Douglas Aircraft, General Electric and several other firms. The companies suggested that such a programme would cost around $20 billion. In the end, the military and Nasa spent possibly as much as $50 billion to land a man on the Moon. Much of that, of course, was channelled to the manufacturers of the spacecraft, launch vehicles and auxiliary equipment. A 1960 guide to Investing in American Industries phlegmatically notes that since the early 1930s, ‘no airliner programme of any size has been developed without some military aid’ and that ‘aid from the Pentagon will remain prerequisite to major advances in the state of the art of commercial air transport.’

The justification usually made for such expenditure is that the $50 billion funded inventions which had more widespread uses: LED lamps, artificial limbs, even the nutrient additives in baby formula. Perhaps the biggest beneficiaries were electronics firms. The microprocessors sold by Intel were invented in conjunction with the Pentagon. Programming languages like COBOL and FORTRAN were written in labs funded by the Pentagon. The internet emerged out of ARPANET (Advanced Research Projects Agency Network), created in 1969 to connect UCLA, UC Santa Barbara, Stanford and the University of Utah to the Pentagon. Sensors that recognise motion (and turn on lights or a tap) or heat (and turn off the oven or your smartphone) were developed during the Second World War and tested in Vietnam to track and bomb guerrillas on the Ho Chi Minh Trail (though the guerrillas easily confounded them with decoys).

In the US the debate over lavish defence spending takes place between conservatives who want to concentrate defence spending in a few firms (preferably run by their allies) and liberal hawks who want to spread the munificence far and wide and use it as an engine of growth. This narrow set of options has shaped the US defence apparatus all the way from The Federalist Papers to DOGE. Whatever the domestic repercussions, the US imperial machine gets the money it needs. Franklin Roosevelt’s secretary of war, Henry Stimson, was sanguine about this: ‘If you are going to try to go to war, or to prepare for war, in a capitalist country, you have got to let business make money out of the process or business won’t work.’ The journalist I.F. Stone put it differently: the firms getting rich on public funding were ‘collective property but under private control’.

According to the historian Ira Katznelson, between 1808 and 1848 federal military spending accounted ‘for at least 72 per cent of the total each year and sometimes up to 94 per cent of federal spending’. After 1848, as the federal government expanded its civilian bureaucracy, the military still accounted for around half of all federal expenditure. In the years between the Civil War and the First World War, even as the national debt and government deficit soared, defence expenditure remained steady at around a quarter of the budget, before rising again to 80 per cent during the Second World War.

In 1969, as the US air force rained defoliants, napalm and bombs on Indochina and the US army hunted guerrillas in the jungles and river deltas of Vietnam, the former commander in chief of US forces there, General William Westmoreland, had a fantasy:

On the battlefield of the future, enemy forces will be located, tracked and targeted almost instantaneously through the use of data links, computer assisted intelligence evaluation, and automated fire control … Today, machines and technology are permitting economy of manpower on the battlefield … But the future offers even more possibilities for economy … With co-operative effort, no more than ten years should separate us from the automated battlefield.

Fifty-six years later, in March this year, the Atlantic Council published a report with a crude AI-generated cover image of soldiers seated at computers with a giant Reaper drone hovering above. The Final Report of the Commission on Software-Defined Warfare recommends that the Pentagon ‘rapidly transform from a hardware-centric organisation reliant on Industrial Age practices and legacy software to a software-centric one more prepared to meet the demands of deterring and combating Digital Age threats’. The recommendations are just what the Defence Advanced Research Projects Agency (DARPA) has been waiting to hear for six decades.

DARPA was originally managed by a relatively small group of scientists and engineers who guided scientific research in top-tier universities and industry labs, and provided long-term research funding. The agency, which had already spawned the internet, was also an incubator for the development of supercomputers, microprocessors and artificial intelligence (or in the words of J.C.R. Licklider, who worked there, ‘man-computer symbiosis’), as well as the Strategic Computing Initiative (SCI) of the Reagan years, out of which battlefield automation, drone technology, advanced command and control systems and data-mining and intelligence analysis projects were born. In The Closed World (1996), a magisterial history of Cold War computing in the US, Paul Edwards argued that to make the SCI’s extravagant budget palatable to the public, DARPA touted the commercial applications of such innovations and ‘frequently gave Japanese competition equal billing with military need in promoting the plan’.

The Pentagon’s plans​ for a ‘revolution in military affairs’ – the introduction of automation and information technology into all aspects of warfighting – weren’t revolutionary enough for Silicon Valley technology firms and their venture capitalist patrons. In 2016, Thiel and Palantir successfully sued the army over its plans to develop its own in-house data-mining and intelligence analysis system rather than use Palantir’s commercially available software. Thiel was following in the footsteps of Elon Musk, whose SpaceX, using the same attorney, had sued the US air force for not awarding it the contract to launch military satellites.

Shah and Kirchhoff of the Defence Innovation Unit found Palantir and SpaceX’s legal actions useful: ‘Little by little, the lawsuits were forcing the Pentagon to rethink its acquisition process – in part by shaming the DoD and exposing to the public its outdated, oligopolistic methods.’ Their book names another start-up, Anduril, as one of the firms that benefited from these lawsuits. Anduril – named after Aragorn’s sword in The Lord of the Rings; the sword is also known as the Flame of the West – was set up by Trae Stephens, a former army intelligence officer and Palantir alumnus, and Palmer Luckey, a young inventor who had sold his virtual reality headset to Meta for $2 billion. Stephens and Luckey, who planned to produce software to control swarms of drones, received advice from the Defence Innovation Unit and funding from Andreessen Horowitz, one of Silicon Valley’s biggest venture capital firms. One of Anduril’s products is a miniature stealth drone that senses motion and is used overseas by the US military, and by the Departments of Homeland Security and Customs and Border Protection inside the US and along its borders.

Although the billionaire start-up founders and defence tech venture capitalists like to think of themselves as scrappy underdogs fighting the ogres of congressional oversight and unhelpful Pentagon bureaucrats, they have already shifted the patterns of technology ownership not only in commercial spaces, but in the military itself. Instead of the Pentagon – or any other public or private institution – owning the software they pay for, corporations now turn their products into ‘services’. They retain control of these services through a subscription model which ensures continuous rent extraction and the ability of the corporations not only to update or fix the software remotely (and get paid handsomely for it) but also to switch it off at source (as Musk briefly removed Ukraine’s access to his Starlink satellites in 2023).

Musk has since muddied the waters further. The White House has become an advertising backdrop for Tesla, with Trump posing beside the cars; the State Department planned to procure armoured Teslas worth $400 million; and as part of Trump’s tariff and trade orders, Marco Rubio’s diplomats have been badgering other countries to lower their regulatory thresholds for satellite communications. Starlink is named in the diplomatic cables.

Palantir provides a platform for start-ups to access military data useful for pitching projects to the Pentagon. It also offers companies the chance to ‘build their software atop Apollo and Rubix, two already accredited platforms built by Palantir’, presumably in order to guarantee deeper defence dependency on its services. Andreessen Horowitz has a web page dedicated to ‘DoD Contracting for Startups 101’. Pitching to the Pentagon requires patience, it warns, but ‘defence contracts offer not just funding, but also long-term sustainment.’ Marc Andreessen is the author of the self-aggrandising ‘Techno-Optimist Manifesto’, which includes a list of enemies (such as universities, the woke mind virus, degrowth, sustainability and, inexplicably, ‘Thomas Sowell’s Unconstrained Vision, Alexander Kojève’s Universal and Homogeneous State, Thomas More’s Utopia’) and demands a return to the origins of Western civilisation, which was built, he claims, on technology.

Karp’s book is a call for Silicon Valley to heed its ‘affirmative obligation to participate in the defence of the nation and the articulation of a national project’. He criticises Silicon Valley’s focus on advertising and customer service, the ‘defanging of Germany’ and the ‘highly theatrical commitment to Japanese pacifism’. The US, he warns, has not come to terms with ‘the rise of an assertive and capable China as well as a newly ambitious Russia’. The correct response to this state of affairs is, of course, to invest in the technologies that Palantir, Anduril and others provide.

The United States was born in war and has waged a war of some sort in every year of its existence. Silicon Valley knows that war is good for business. And many of its most powerful people want us to stop worrying about frivolities like ethics or ecology and love the bomb. In Male Fantasies (1977), a psychoanalytic reading of the rise of fascism in Germany, Klaus Theweleit described the ecstatic commitment of the Freikorps to their mission. What a soldier demands is ‘a war in which he experiences the whole of his being and his future potential. In and across the machines with which he sets off to war, the man consolidates his existence as man; it may be in war that he becomes a man in the first place.’ For the armchair techno-warriors of Silicon Valley, the barbarians at the gate are a useful solution.