The socio-legal relevance of artificial intelligence and the normativity of algorithmic design
Activity: Talk or presentation › Invited talk
|Title||The socio-legal relevance of artificial intelligence and the normativity of algorithmic design|
|Person and role|
|Description||There has been much talk of artificial intelligence (AI) lately. But, as AI is becoming more commonplace on consumer markets, how do you ensure fairness, accountability and transparency for technologies that are autonomous and self-learning?
There is little doubt that AI, and particularly machine learning utilising large amounts of data even in real-time, can make many sorts of processes more efficient and individually relevant for consumers – be it in autonomous cars, smart homes, medical diagnosis, risk assessments or distribution of welfare. There has however been an increasing debate around risks of autonomous systems reproducing or even amplifying social bias, such as inequalities in gender, socio-economic groups or ethnicities. So, how to detect unfair practices and deal with accountability for autonomous systems and AI is of increasing legal concern.
This talk addresses what AI is and increasingly means for consumers in everyday life. It addresses questions of how to understand transparency in this context and how to understand issues of accountability for autonomous and algorithmic processes, for example by addressing in what way data-driven design is a normative practice generating responsibilities.
A challenge I see for lawyers in the context of AI relates to methods required to discover unintended or bad consequences. It probably means that lawyers and technologists need to work more closely together.
2018 Oct 19
|Research areas and keywords||
|Period||2018/10/19 → 2018/10/19|
|Web address (URL)|
|Location||Flemish Bar Association|
|Degree of recognition||International event|
The Swedish Retail and Wholesale Council (Handelsrådet)
2018/09/01 → 2020/08/31
Related research output
Research output: Contribution to specialist publication or newspaper › Newspaper article
Research output: Contribution to journal › Article