What form of governance is necessary for AI applications? What role should DPAs play in their supervision? This panel will explore the challenges raised by AI for DPAs, the new instruments, tools and frameworks DPAs should develop to tackle these challenges and the initiatives already in place. What are the opportunities and limits of regulatory sandboxes or AI labelling? What role for ethics in the supervision of AI? Is the creation of EU “data lakes” for research and training of algorithms a way forward?
Building on the outcome of the first discussions that will take place in a World Café setting on Tuesday 21 January, the panel will focus on how data protection authorities can confront the multi-faceted human rights impact of AI technologies. In the public authority context, the panel will debate on how to uphold important legal concepts such as the presumption of innocence, as well as due process and non-discriminatory treatment. Since AI is not a stand-alone technology, the panel will also discuss the difficulties to anticipate or even identify the risks raised by these technologies, including the ones associated to the cumulative effects that propagate small flaws and inaccuracies in a disproportionate and uncontrolled way.