BEYOND ETHICS WASHING:IMPACT ASSESSMENTS, AUDITS, AND OVERSIGHT FOR AI

CPDP 2023 - DAY 3
AREA 42 GRAND - SESSION 1
Organised by Helsinki Institute for Social Science and Humanities,
University of Helsinki (FI)
Moderator Mirko Tobias Schäfer, Utrecht University/University of
Helsinki (NL/FI)
Speakers Paul Nemitz, European Commission (EU); Iris Muis,
Utrecht University Data School (NL); David Graus, Randstad (NL);
Maria Koomen, Open Governance Network for Europe (BE)
A plethora of ethics manifestos, guidelines and frameworks calls for
responsible AI and data practices. Legislation is under way to regulate
AI practices. But how to effectively close the gap towards practical application? How can organisations implement practices that stimulate
responsible application of AI systems, and how can increasingly digitized democratic societies establish necessary checks and balances?
Looking further than the good intentions of ethics guidelines, this panel
discusses which best practices are most effective to align the design
and use of AI systems with the values of our open and democratic societies. This panel investigates how practical approaches, such as impact assessments and audits, and the role of oversight bodies help to
establish responsible and safe uses of AI and big data practices and can
constitute accountability.
• Should impact assessments be mandatory?
• Who is responsible for oversight and enforcement?
• Can we hold algorithms accountable?
• What are the limits to good governance of AI?