Digital Trust & Security Guest Seminar Series: Dr Andreas Theodorou
|Starts:||12:00 9 Feb 2022|
|Ends:||13:15 9 Feb 2022|
|What is it:||Seminar|
|Who is it for:||University staff, External researchers, Alumni, Current University students|
At this Digital Trust & Security Seminar, we will be joined by Dr Andreas Theodorou who will be discussing meaningful control of AI.
Speaker: Dr Andreas Theodorou is a Forskare (Research Fellow) in the Department of Computing Sciences at Umeå University, Sweden.
Talk title: Accountability, Responsibility, and Transparency for Human Control
Summary: The integration of Artificial Intelligence (AI) in our societies comes with the increased public demand for human control. Maintaining effective control over such systems is an ongoing socio-technical challenge that has sparked global discussions and recommendations in the form of soft policy documents, i.e. guidelines and standards, and the creation of control frameworks for human oversight. Such guidelines rely heavily on high-level statements that provide no clear criteria for system assessment. Furthermore, proposed control framework put the onus of responsibility exclusively to the human end user; away from the developers and deployers of those systems. In this online talk, I discuss the concept of meaningful human control which extends beyond mere oversight by including design and governance layers. I propose how the combination of context-specific AI assessment and variable autonomy—dynamically adjustable levels of autonomy—can ensure such control over an artefact and satisfy all three core values commonly advocated in ethical guidelines: accountability, responsibility, and transparency.
Speaker bio: Dr Andreas Theodorou is a forskare (tenured research fellow) in Responsible Artificial Intelligence (AI) at Umeå University. He is working at producing the techniques, processes, and tools needed to understand and mitigate the socio-economic, legal, and other ethical issues and challenges that arise from integrating AI into our societies. Andreas is currently the Primary Investigator for the RAIN project (KAW funding), which is looking into context-specific AI Assessment. In parallel to his research, Andreas is an active member of various AI Governance initiatives (e.g. IEEE SA’ P7001 series, ISO JTC1/42, BSI ART/1) and of AI Ethics boards (e.g. ROXANNE project). Prior to his current post, Dr Theodorou has been affiliated with the University of Bath (UK), the Georgia Institute of Technology (USA), and the University of Surrey (UK).
Travel and Contact Information