Love in the time of Palantir

Love in the Data: Palantir, Dating, and Surveillance

Episode guide Published Planet Money 29 min

概览

This Planet Money episode uses a satirical New York singles event, “Singles Dating Matching Using Palantir Data,” to explore what people fear and misunderstand about Palantir, data surveillance, and modern data-matching tools.

The episode distinguishes between Palantir’s public image and its stated business model: Palantir does not own, buy, or sell data, according to reporter Mike Steinberger, but provides software that helps governments and companies connect messy data sets and make them usable.

Through the dating event, the show demonstrates how much information can be assembled from names, phone numbers, public records, facial recognition, leaked data, and other sources. The results are uneven: some profiles are wrong, while others reveal unexpectedly specific details that unsettle attendees.

The broader conclusion is that the danger may not be Palantir personally holding everyone’s data, but that powerful software can make already-available or government-held data far easier to combine, search, and act on.

分段落总结

[00:28] A Dating Event Built Around Palantir’s Image

[事实] The episode opens with an Instagram ad for a New York event called “Singles Dating Matching Using Palantir Data.”

[事实] The ad claimed attendees might find love using technology associated with tracking Osama bin Laden, though the episode notes that Palantir’s role in that has never been confirmed.

[事实] Around 60 people bought $20 tickets, and most gave the organizer their names and phone numbers.

[推测] The event works as both comedy and a public experiment about what people imagine data companies can know about them.

[02:12] Attendees Want to Know What Data Reveals

[事实] Several attendees say they came because they were curious about what could be found about them from limited personal information.

[事实] Some attendees describe the event as a privacy-awareness exercise, saying people should know what others can discover about them.

[事实] One attendee, Jeannie Chan, says her common name and a same-name professional in New York might make her harder to identify accurately.

[推测] The crowd’s curiosity reflects a broader uncertainty about the difference between public data, brokered data, government data, and Palantir’s actual role.

[05:19] James Steinberg’s Satirical Tech Event

[事实] Organizer James Steinberg has run other strange tech-themed events and describes himself as a tech person who throws comedic, weird events.

[事实] James has founded startups and worked as a cybersecurity CTO, so he has experience with online privacy.

[事实] He says he is not using Palantir data or Palantir software for the dating event.

[事实] The episode notes Palantir’s controversial work with ICE, the Department of Defense, and uses in Gaza, Ukraine, and Iran.

[推测] The satire lands because Palantir’s brand is already strongly associated with surveillance, military applications, and secrecy.

[07:09] What Palantir Actually Does

[事实] Mike Steinberger says Palantir does not own data, buy data, or sell data.

[事实] Palantir works with agencies and organizations including the CIA, Department of Homeland Security, ICE, Department of Defense, FBI, IRS, Department of Justice, foreign governments, and businesses.

[事实] Palantir’s forward deployed engineers may see client data when working closely with a customer, but the company’s software generally helps customers use their own data.

[事实] The software logs actions performed on the platform, which the episode describes as one privacy-related guardrail.

[08:40] The Airbus Example

[事实] Palantir worked with Airbus when the company was having production problems with the A350 airplane.

[事实] Airbus had difficulty seeing its entire production line because assembly involved eight plants across four countries.

[事实] Palantir software created a real-time visual representation of the A350 production line and helped identify breakdowns faster.

[事实] The episode says Airbus reduced the time to find and fix production glitches by about a week.

[推测] This example shows why Palantir can look mundane in business use while still being powerful when applied to sensitive personal or government data.

[10:20] Can Palantir Logic Make Love Matches?

[事实] Mike Steinberger says Palantir’s software is flexible and could theoretically be used for many purposes, including matchmaking.

[事实] At the event, James tells attendees he has never worked for Palantir, is not using Palantir software, and has not been endorsed by the company.

[事实] James says Palantir sent him a cease-and-desist email telling him not to use the word Palantir.

[推测] The legal tension around the event reinforces how valuable and sensitive Palantir’s brand identity has become.

[12:02] How the Dating Profiles Were Built

[事实] James says he spent about 50 hours using attendees’ names and phone numbers to build dating profiles with Palantir-inspired methods.

[事实] He used data enrichment, facial recognition software, reverse phone lookup, bounty hunter databases, court records, and leaked data from the dark web.

[事实] James is described as a licensed bounty hunter.

[事实] He presents unusually specific facts about attendees, including details about Becca attending a wedding in Cleveland.

[推测] The segment shows that invasive-feeling profiles can be built without access to a secret corporate database.

[13:05] Jeannie’s Profile Shows the Limits of Matching

[事实] Jeannie and Zach are called on stage as a potential match.

[事实] The event displays information claiming Jeannie had some connection to Indonesia’s Ministry of Agriculture, but Jeannie says that was not her.

[事实] Photos shown as Jeannie are also not her.

[事实] The episode notes Jeannie had predicted this problem because someone with the same name, job, and city exists.

[事实] The host says facial recognition software is notoriously poor at recognizing people of color.

[推测] Jeannie’s case shows that data-matching systems can be invasive even when they are wrong.

[15:30] Zach’s Profile Is More Accurate and More Disturbing

[事实] Zach’s profile correctly identifies that he is a machine learning engineer and worked for Nike.

[事实] The profile also identifies that he built an online tool for the Bernie Sanders 2020 campaign, marched in an anti-ICE demonstration, and spent a week in Goa.

[事实] Zach says he had not posted about the Goa trip or the anti-ICE protest.

[事实] Photos of Zach at a protest and a concert are shown, even though he did not know they had been taken.

[事实] Zach says he had not thought of himself as consenting to being searchable by his face.

[推测] Zach’s reaction gives the episode its clearest example of how public life can become searchable without meaningful consent.

[18:02] The Event Becomes a Privacy Debrief

[事实] After more potential couples and mixed-accuracy profiles, attendees go to a bar to mingle and process the event.

[事实] The episode then shifts from the dating experiment to how Palantir tools can help the government assemble files on millions of people.

[推测] The dating-event structure makes the surveillance discussion more concrete by first showing it at an individual, social scale.

[19:53] Palantir’s Post-9/11 Government Origins

[事实] Mike Steinberger says Palantir was founded after 9/11 and pitched itself to the CIA as a way to connect data that might help prevent future attacks.

[事实] Palantir received a CIA contract in 2005 and embedded with analysts to build software for tracking and creating files on terrorism suspects.

[事实] The episode explains that government data had often been intentionally siloed because of past abuses.

[事实] Examples include Census Bureau data being used during Japanese American incarceration and Nixon using FBI data against enemies.

[事实] The Privacy Act of 1974 restricted government agencies from repurposing data collected for one reason without consent.

[22:11] Data Silos, ICE, and Larger Government Files

[事实] The episode says President Trump signed an executive order called “Stopping Waste, Fraud and Abuse by Eliminating Information Silos.”

[事实] The episode says there was a memorandum of understanding between ICE and the IRS because ICE wanted taxpayer information.

[事实] Mike Steinberger says ICE can now access 30 to 40 different data sets.

[事实] Palantir software can help enrich separate data sets, find patterns, and create more detailed files about individuals and their networks.

[事实] Steinberger says Palantir appears to be the technological backbone for a large ICE surveillance apparatus.

[23:39] Speed, Scale, and Civil-Liberties Concerns

[事实] The episode cites 404 Media reporting that Palantir software helped ICE locate people in minutes instead of hours.

[事实] The episode says the database apparently contains 20 million names.

[事实] The host notes that this number is larger than the estimated undocumented population in the United States.

[事实] Mike Steinberger says Palantir does not have the information itself, but makes ICE’s job easier.

[推测] The concern is that tools built for one target population can later be directed at other groups.

[24:32] Palantir’s Response and Oversight Problem

[事实] Palantir declined to be interviewed for the episode.

[事实] The episode says Palantir has previously responded to critics by saying it is not building a master surveillance database for the government.

[事实] Mike Steinberger says Palantir does not ultimately control what customers do with its technology.

[事实] ICE did not respond to the show’s request for comment.

[推测] The episode frames customer control as a major unresolved issue: even if the software company does not own the data, its tools can still enable powerful state action.

[25:25] Did the Dating Event Work?

[事实] As of September 2026, Palantir had not sued James or others involved in the dating event.

[事实] James booked two more similar events.

[事实] Zach says he met someone he was interested in seeing again and had a drinks date planned.

[事实] James says at least six phone numbers were exchanged, four dates were made, and one couple began a date immediately after the event.

[推测] The event succeeded as matchmaking partly because public exposure made it easier for interested people to approach each other afterward.

播客点评/总结

[推测] The episode’s strength is its structure: it uses a strange dating event to make abstract questions about data surveillance feel immediate, awkward, and personal.

[推测] Its most valuable point is the distinction between owning data and operationalizing data. The episode argues that the real power lies in connecting scattered information quickly and making it actionable.

[推测] A limitation is that Palantir and ICE did not participate in interviews, so the episode relies on reporting, public statements, outside experts, and the dating-event experiment rather than direct answers from those institutions.

[推测] This episode is best suited for listeners interested in technology, privacy, government surveillance, data brokerage, AI-adjacent culture, and the social consequences of making people searchable.