Cornell University
Search:
more options

Levy, Barocas Paper Makes Notable Must-Read List

A crucial paper from Info Sci professors Karen Levy and Solon Barocas that explores user bias on online platforms has been named one of six must-read papers of 2017 by a leading privacy-law group. 

Designing Against Discrimination in Online Markets” centers in on 10 categories of design and policy decisions that perpetuate, exacerbate or alleviate discrimination among users of popular online platforms, from dating apps and Craigslist to Uber and AirBnb. It was chosen by Future of Privacy Forum (FPF) as part of the non-profit’s annual Privacy Papers for Policymakers Award, which recognizes the latest, ground-breaking privacy research meant to inform policymakers on privacy issues and offer real-world solutions. As part of their paper's selection, Levy and Barocas will be invited to the U.S. Senate in February to present their work to lawmakers, academics and privacy professionals.

In their notable paper, the Tech-Law/Policy experts write: “The early web was marked by considerable uncertainty as to the reliability of the person on the other side of some exchange. Today’s web is dominated by platforms that employ a diverse set of techniques to relieve users of such anxieties—providing assurances that can go far beyond what people might glean from in-person interactions. 

“In adopting these techniques, however, platforms have begun to exhibit the sorts of worrisome dynamics that are common in face-to-face encounters.” 

Details revealing gender, race, and ethnicity open the door to discrimination, either consciously or unconsciously, the pair writes, and design and policy choices make user-to-user interaction on these platforms more or less conducive to discrimination. 

“Platforms cannot divest themselves of this power,” they write. “Even choices made without explicit regard for discrimination can affect how vulnerable users are to bias.” 

This is Barocas’s second consecutive year on FPF’s annual list. A paper he coauthored “Accountable Algorithms” was among the 2016 year-end picks.