In Q Tel, the Central Intelligence Agency's investment fund, gave Palantir a generation long head start inside US intelligence procurement, which is now visible in NHS patient data and UK defense contracts.
Palantir's grip on Western state infrastructure was not built in the marketplace. It was seeded by the CIA's venture arm, In-Q-Tel, two decades before any of the current controversies. That is why the news in 2026 looks the way it does.
In May, UK Members of Parliament called Palantir's access to identifiable NHS England patient data "dangerous," and a separate Guardian report noted that co-founder and chief executive Alex Karp's manifesto-style post on X was dismissed in UK political coverage as "the ramblings of a supervillain." The BBC's coverage of the same row sits inside the same frame. Read those together as a cycle, and they look like familiar political outrage against a controversial vendor. Read them as the visible surface of a longer arrangement, and they look like something else.
That arrangement has a paper trail. In-Q-Tel, the Central Intelligence Agency's venture-capital arm, was an early institutional investor in Palantir, according to Fortune's 2025 profile of the fund and its portfolio companies. The relationship runs back to Palantir's founding years, when the company existed in significant part to commercialize tools the US intelligence community wanted. Once a national intelligence customer has spent roughly two decades co-developing a vendor's software around its own operational assumptions, the cost of switching is not a budget line. It is a working culture. That is path dependency, and it is what Palantir has spent the last decade exporting to allied governments.
This is also why the techno-oligarchy framing, often dismissed as slogan, works as mechanism in this case. The Fair Observer analysis that crystallised the current debate draws on the political economist Kean Birch's concept of the "totalitarian corporation," a firm whose internal coordination is tight enough, and whose entanglements with state power are deep enough, that it competes with states rather than only with other companies. Birch coined the term in 2007 and has continued to develop it. The essay applies the framework to a small cluster of AI, surveillance, and data-mining firms entangled with Western governments. Palantir is the test case because the entanglement is documented, not alleged.
The political vocabulary in Westminster is, in 2026, catching up to a procurement relationship that has been settled since the early 2000s. The watch item is concrete. If NHS England's data architecture and the UK's defense procurement stack remain Palantir-dependent after this controversy and after the next procurement cycle, then the dependency was never really in question. The recent headlines will have functioned as a release valve rather than a correction.