Photograph Source: UK Government – Deputy Prime Minister Oliver Dowden attends AI Summit – CC BY 2.0

Palantir CEO Alex Karp’s description of the company’s quarterly earnings in August 2025 as “once in a generation” wasn’t far-fetched. It surpassed $1 billion in quarterly revenue for the first time, after only turning profitable in 2022. Its surge has made it an outlier: still too small by revenue to crack the Fortune 500, its roughly $400 billion market value by December (more than double what it was in January) places it among the 25 most valuable companies in the world.

Palantir’s ascent over the last couple of years comes after two decades of operating largely outside the public spotlight, at least compared to Big Tech giants or Elon Musk’s companies. Founded in 2003, it went public only in 2020 and has fewer than 4,000 employees. It runs no public advertising and sells no consumer-facing products, focusing entirely on developing software platforms for government agencies, companies, and nonprofits.

Part of Palantir’s success stems from what amounts to a monopoly. In his 2014 book Zero to One and follow-up essay “Competition Is for Losers,” co-founder Peter Thiel argued that the most successful companies are “creative monopolies,” and are so good at what they do that no real substitute exists. Citing Google’s search engine, Thiel wrote that monopoly profits let companies invest more in workers, products, and long-term innovation, while disrupting stagnant incumbents.

Understanding Palantir’s business monopoly is difficult because both the firm and many of its clients keep its platforms largely out of view. Named after the seeing stones in The Lord of the Rings, Palantir grew out of PayPal’s fraud detection project. Its operating systems now combine data from communications, supply chains, operations, and other sources to reveal patterns and support faster decision-making. Over time, Palantir has layered these tools into integrated management systems, with advisers often deployed to simplify adoption.

At the center is its “Ontology,” a central brain that links an organization’s data to show how everything fits together, “transforming raw data into valuable knowledge.” Originally a proto-AI system, it has been enhanced since 2023 with Palantir’s Artificial Intelligence Platform (AIP), which uses large language models, generative AI, and coordinated AI agents that work together. This setup reduces scaling constraints and enables a more “plug-and-play” approach that can be managed remotely.

Clients can build comparable systems in-house, but doing so is slow, expensive, and often difficult to maintain. As the first mover, Palantir can deploy advanced, customizable platforms in weeks, and the combination of performance, convenience, and high switching costs tends to lock customers in.

Developing the Domestic Engine

Palantir’s platforms have been used in the U.S. for decades, initially by agencies seeking advanced planning and security tools after 9/11. Early access to unique government datasets trained its systems on everything from modern databases to decades-old legacy software. Its tools can integrate radar, satellite imagery, photographs, communications records, human intelligence, and more. Palantir’s

© CounterPunch