A recent allegation has been levied on Facebook stating that the algorithm used for delivery of ads to the user is discriminatory and is based strongly on stereotypes related to gender and ethnicity despite the advertisers claiming that their parameters are inclusive invariably.
The study has been called Discrimination through Optimization and was conducted by a bunch of researchers from the University of Southern California and Upturn and the Northeastern University who claimed that the ad delivery can lead to skewed distribution. The paper is not officially out yet as they are still deciding the ideal forum to do so. The study comes in the wake of a recent charge by the Department of Housing and Urban Development under the administration of Donald Trump about the advertising practices used for housing and seeking damages for parties who have been a victim of targeted policies.
Surprisingly, landlords and employers could actually limit their prospective audience until a while ago based on their race, gender, and ethnic origin. Another implication of the study is the fact that how less attention had been given to this kind of discrimination earlier. Financial and market optimization has been touted as the reason that this skewed presentation has taken place and is instrumental in making the predictions to provide advertisements to users. The budget and the content of the ad also played a significant role but the presence of racial and gender implications were also noticed. This was the case even when the advertiser decided to be neutral in their selection. There is a chance that the advertisers are not even aware of what is happening.
What this study does is expose the problems in online advertising. A spokesperson from Facebook has released a statement stating that they are strongly against any form of discrimination and have made appropriate changes to the advertisement platform accordingly.