He’s not the only one.

Photographer: Drew Angerer/Getty Images

How Data Geeks Could Enable Mass Deportation

Elaine Ou is a blockchain engineer at Global Financial Access, a financial technology company in San Francisco. Previously she was a lecturer in the electrical and information engineering department at the University of Sydney.
Read More.
a | A

Silicon Valley tech workers are worried that Palantir Technologies, a company co-founded by Donald Trump adviser Peter Thiel, could end up helping the president-elect crack down on immigrants based on their race, religion or national origin. They should also recognize the role that they -- and the companies they work for -- have played in making such a scenario possible.

This week, the workers staged a protest outside Palantir’s headquarters in Palo Alto. Their demand: that the company, which sells data-mining technology to the government and other customers, prevent the federal government from using the contracted software to create a Muslim database or facilitate mass deportations. Their concerns are focused on two systems: Falcon and the Analytical Framework for Intelligence, both of which the Department of Homeland Security uses to profile individuals -- including such information as social and family relationships, travel histories and addresses.

Yet Palantir’s analysis tools are not terribly different from what these folks work to create every day. Technology companies already build databases that document people’s ethnicityreligion and other personal information that allows them to target ads. Major networks such as Google and Facebook can infer detailed demographic information by tracking users all over the internet with cookies and device identifiers, and form partnerships with data brokers that provide the personal information needed to connect that online activity to specific individuals. One company engaged in “identity resolution,” LiveRamp, says it can recognize 98 percent of U.S. adults based on online identifiers.

It’s not clear how much the government uses such commercial data -- though it certainly has the capacity to do so. The Department of Homeland Security states that Falcon accesses commercially available databases, though it doesn’t specify which ones. Many commercial data brokers sell their services to government agencies: One of the largest, Acxiom (owner of LiveRamp), has a long history of working with the government and was involved in a post-9/11 Homeland Security initiative to screen air passengers that raised concerns about the leakage of personal data. Acxiom’s partners include Google, Facebook, Twitter, Yahoo, Pinterest, Foursquare and just about every other company that serves up ads.

If tech companies are serious about preventing undue surveillance and profiling, they should stop contributing to the information that could enable it. That means ending partnerships with data brokers, limiting their own data collection and not doing creepy things like tracking users’ physical location to see which stores they visit. While I appreciate these companies’ great efforts to show me relevant advertising, I would also appreciate it if they didn’t bring my Safeway shopping history into the picture.

The best way to protect data from being used for undesirable ends is to refrain from storing it in the first place. To resist Trump’s proposed deportation plans, for example, the New York City mayor’s office announced that it would no longer retain personal records for immigrants who apply for IDNYC identification cards. No search warrant or court order can be used to obtain information that does not exist.

There are plenty of reasons to protest Trump’s immigration policies. But placing the blame on a single technology provider is misguided. Palantir doesn’t have a monopoly on data analysis software. Even if every technology company in the country refused to build the government’s immigration databases, dozens of international providers would love to land the contract.

If Trump wants to target millions of people for deportation, a lack of know-how is not going stand in his way. A lack of data, by contrast, could at least slow him down.

  1. Online databases are supplemented with information collected from store loyalty cards, public voting records, and salary and pay stubs to form comprehensive user histories.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Elaine Ou at elaine@globalfinancialaccess.com

To contact the editor responsible for this story:
Mark Whitehouse at mwhitehouse1@bloomberg.net