Data Broker Inferences Can be Wildly Inaccurate

Data brokers are already under scrutiny concerning their core business of collecting and selling personal information. Now, data broker inferences are being criticized for inaccuracies.

Suzanne Smalley, reporting for The Record, says data brokers’ clients include advertisers, marketers, insurance companies, researchers, and often law enforcement and government.

According to Smalley, some experts warn that this kind of “inaccurate trash” can affect not only the ads people see, but also their financial affairs, their insurability, their healthcare, and the opportunities they are or are not offered online. 

“Insurance companies, for example, regularly buy personal data to inform insurance pricing, so if data in there is inferred or portrayed incorrectly that could result in an erroneous spike in someone’s rate,” says one industry observer interviewed by Smalley. 

Another source, the president of a data media and services company, tells Smalley that he’s seen himself inferred to be a heavy drinker, a smoker, and a participant in extreme sports. In fact, he is none of those things. He says he’s also been listed as living in the wrong state, in one case at an address that he determined to be a vacant block in New York. 

Problems have also been alleged in cases where data broker inferences are arguably accurate. A report from the Connecticut attorney general’s office determined that a national cremation service was targeting ads to state residents who had recently finished chemotherapy.

The faulty inference phenomenon has led to the emergence of another industry, one that purports to help data brokers improve their processes. Smalley interviews a principal at one such company, which lists among its clients third-party data brokers Experian, Epsilon, and TransUnion.