Deutsch Intern
Würzburg Centre for Social Implications of Artificial Intelligence (SOCAI)

Subproject 3: state liability and risk management in automation

The use of AI -enabled tools, in particular machine learning, in public administrative procedures brings about uncertainty in terms of attribution of liability to state and its civil service as a result of the potential of unforeseen errors that could be generated. In this context, the position of individual public administrators and the degree and scope of their involvement within a given course of action would require further analysis.  Here questions related to causality and the attribution of responsibility to the state as a result of an error by AI-enable tools may result in the exoneration of personal liability of individual public administrators.