Thursday, February 23, 2017
How Peter Thiel’s Palantir Helped the NSA Spy on the Whole World
Palantir has never masked its ambitions, in particular the desire to sell its services to the U.S. government — the CIA itself was an early investor in the startup through In-Q-Tel, the agency’s venture capital branch. But Palantir refuses to discuss or even name its government clientele, despite landing “at least $1.2 billion” in federal contracts since 2009, according to an August 2016 report in Politico. The company was last valued at $20 billion and is expected to pursue an IPO in the near future. In a 2012 interview with TechCrunch, while boasting of ties to the intelligence community, Karp said nondisclosure contracts prevent him from speaking about Palantir’s government work.
“Palantir” is generally used interchangeably to refer to both Thiel and Karp’s company and the software that company creates. Its two main products are Palantir Gotham and Palantir Metropolis, more geeky winks from a company whose Tolkien namesake is a type of magical sphere used by the evil lord Sauron to surveil, trick, and threaten his enemies across Middle Earth. While Palantir Metropolis is pegged to quantitative analysis for Wall Street banks and hedge funds, Gotham (formerly Palantir Government) is designed for the needs of intelligence, law enforcement, and homeland security customers. Gotham works by importing large reams of “structured” data (like spreadsheets) and “unstructured” data (like images) into one centralized database, where all of the information can be visualized and analyzed in one workspace. For example, a 2010 demo showed how Palantir Government could be used to chart the flow of weapons throughout the Middle East by importing disparate data sources like equipment lot numbers, manufacturer data, and the locations of Hezbollah training camps. Palantir’s chief appeal is that it’s not designed to do any single thing in particular, but is flexible and powerful enough to accommodate the requirements of any organization that needs to process large amounts of both personal and abstract data.