Roundtables: Why It’s So Hard to Make Welfare AI Fair
1 min read
Summary
In the quest to create a fair and discriminatory-free welfare system, Amsterdam’s attempt to assess welfare applicants using AI has backfired due to bias still creeping into algorithms.
Despite testing and attempts to eliminate bias, it appears that discriminatory algorithms were built into the welfare system, similar to criminal justice and healthcare systems.
This raises questions about whether AI and algorithms can ever be fair and if so, how?
This topic will be discussed in more detail in a forthcoming panel discussion by MIT Technology Review that invites views from its editor Amanda Silverman, investigative reporter Eileen Guo and Lighthouse Reports investigative reporter Gabriel Geiger.
Links to related readings will also be made available to attendees ahead of the online event.