Marsilius-Seminar: Risk management, Fairness and Transparency for AI
Risk management, Fairness and Transparency for AI
AI as a disruptive technology has recently become the focus of ethico-legal discussions, especially in the context of ChatGTP. The question of what AI can do and where the limits of AI lie is ultimately accompanied by questions of trust in the technology and responsibility for training, using, and evaluating the results of AI systems. The backdrop to this question is generally determined by the overall regulatory framework and the expected changes that AI applications might bring about at the societal level. The seminar will therefore focus on trust, responsibility and risk management, and fairness in the use of AI. The overarching principle that will guide our analysis of these aspects is transparency.
For this reason, one of the questions we will explore is what measures need to be taken to address the lack of standards for verifying AI results. How can AI bias be reduced and fair automated data processing ensured? What ethical governance is necessary for trustworthy AI? How does the societal and sectoral context influence ethical-legal standards of AI development and use?
We would like to shed light on transparency from different perspectives in an interdisciplinary framework to answer the question whether transparency is an unrealistic requirement of data protection law or a feasible condition to guide the development and use of AI. To this end, participants will prepare presentations in an interdisciplinary tandem. The event is aimed at students from all faculties and disciplines; prior knowledge of data protection law is helpful but not mandatory.
11.12.2023 14:00-17:00 BioQuant SR042
05.02.2024, 06.02.2024 08:00-18:00 BioQuant SR042
More Information on LSF and at www.marsilius-studien.uni-hd.de. or contact. marsilius-studien[at]uni-heidelberg.de