Having earned a 9 for his thesis, International and European Law alumnus David Dalenberg then went on to win the 2017-2018 UvA Thesis Prize. His thesis deals with strategies for avoiding discrimination in targeted job advertisements. The 3000 EUR prize was awarded during the UvA University Day.
Each year on the UvA University Day, UvA Thesis Prizes are awarded for the best and most original Master’s theses of the previous year. Not everyone is cut out for this job, though. Students need to submit two letters of recommendation in addition to their list of marks and final mark. David wrote his thesis Preventing discrimination in the automated targeting of job advertisements for the Master’s track Public International Law. Kevin Jon Heller and Frederik Zuiderveen Borgesius penned letters of recommendation.
The thesis examines discrimination caused by the use of artificial intelligence in targeted job advertisements. AI is used more and more frequently to automate decision-making, e.g. for mortgage and other loan approvals, job recruitment, and even for university admissions.
David: “I looked at what happens when AI is used in recruitment procedures, the automated targeting of job advertisements in particular. AI can be used, for instance, to determine whether you will be shown the advertisement for a specific job. Imagine that an ad for a CEO position is only made visible to men – women would then obviously have a poorer chance of success. Previous research has shown that this does actually happen.”
According to Frederik Zuiderveen Borgesius, the thesis reveals a “deep understanding not only of the interplay between artificial intelligence, big data, algorithms and machine learning, but also of EU non-discrimination law. The result was an exceptional thesis.” The thesis has in the meantime been published as a paper in the academic research periodical Computer Law & Security Review.
David thinks his thesis topic bears social relevance due to the fact that so many organisations currently bank on artificial intelligence. “AI and Big Data allow for a greater degree of automatization than ever before but can also inadvertently facilitate discrimination if carelessly programmed software is used. AI-developers need to be more aware of the possible risks engendered by their creations. Based on European non-discrimination law, David has come up with suggestions on how AI should be programmed so as to avoid the risk of bringing about discrimination.
The youth editorial board of Spui25 “Spui in Spe” produced a podcast on David Dalenberg’s thesis.