Document Type
Article
Publication Date
2024
Abstract
(Excerpt)
Since 2018, Uber has submitted applications for numerous patents that use algorithms to “define” safety. These patents “calculate” safety through multiple factors, including crime reports and statistics, news databases, academic databases of reports of violent conflicts in a location, the car’s condition, how often the driver swerves, and “social media.” These machine-learning models attempt to predict “the likelihood that a driver will be involved in dangerous driving or interpersonal conflict.” Drivers are generally outraged by these patents and have commented that these recorded metrics will be “used to manipulate and influence” driver behavior. There is merit to this fear. For example, in one patent application, Uber has associated a lower safety score with drivers who work at night and complete fewer trips. In other words, Uber is evaluating safety for riders through a lens of what it deems safe, which happens to correlate with what may improve its bottom line.
While there are inherent advantages to increased safety measures, the downside of these measures is often overlooked. Transportation Network Companies (TNC) successfully distract drivers and riders from the pervasive monitoring and mandatory job training through the lens of rider safety features. This article will review the way rider safety, driver “flexibility,” and driver monitoring interact, and will ultimately argue that there is probative value in these factors to determine that TNC drivers are employees.
Comments
Second Place Winner - Dr. Emanuel Stein and Kenneth Stein Memorial Law Student Writing Competition
Originally published in NYSBA Labor and Employment Law Journal, Vol. 48, No. 1 (2024). Reprinted with permission from the New York State Bar Association.