Netflix, Inc., which bills itself as the world’s largest online movie rental business, launched its Netflix Prize contest in October 2006. Competitors vying for the $1 million top prize were asked to create algorithms that improved the accuracy of the company’s video-recommendation software. As part of the contest, the company released anonymized data sets containing over 100 million movie ratings submitted by roughly 480,000 subscribers. Sixteen days after the contest launch, a team of university researchers announced that they had successfully identified several Netflix customers from the anonymized data that Netflix released.
On December 17, 2009, plaintiffs based in California, Texas, and Illinois filed a class action complaint against Netflix. See Valdez-Marquez v. Netflix, Inc., 09-cv-5903 (N.D.Cal. Dec. 17, 2009). The plaintiffs argued that customers’ movie selections and reviews constitute highly personal information that the video rental firm should have guarded carefully. They noted that a customer’s film preferences may reveal details about his or her sexuality, religious beliefs, or political opinions. Specifically, the plaintiffs objected to the release of this data in a manner that enabled researchers to match individuals to their movie preferences.
Netflix, in the meantime, had already announced plans for the Netflix Prize 2 competition. For this contest, the company plans to release anonymized subscriber movie selection data, coupled with subscribers’ age, genders, and zip codes.
On March 29, 2010, the district court ordered voluntary dismissal of the un-certified class action complaint, pursuant to a confidential settlement agreement between the parties.
This case demonstrates the complex issues faced by companies who utilize online privacy policies.