One of the most pressing data risks concerns biases and discrimination through data-driven technologies. In December 2020 we organised a critical discussion with experts from diverse fields on the issue.
Data bias and algorithmic discrimination are an important research theme that we approach by addressing the following questions through both theoretical and empirical investigations:
- What is data bias?
- How do algorithms “discriminate” against specific groups of people?
- Where do biases in data-driven technology come from?
- What impact do data bias and algorithmic discrimination have on society
- How can we develop solutions to the problem?
- Who is responsible for data bias and algorithmic discrimination?
We looked into different sectors in society where data bias and algorithmic discrimination have a tangible impact, such as healthcare, education, security, and politics.
While data and artificial intelligence have the potential to offer various benefits to society, there are still challenges for inclusion and fairness. The panelists explored different root causes, effects, and solutions in a critical and engaging debate.
We also turned to the question of responsibility. Participants agreed that hiding behind an algorithmic “blackbox” does not work as human decisions can always be traced back in the use of technology. People can and should be held accountable when data-driven technology has a harmful effect.