Human decision-making is generally related to risk and/or uncertainty. To reduce these, we use past experiences to predict the outcome of our decision. Shared decision-making, especially with AI, is another attempt at minimizing risk, however numerous neuropsychological phenomena like gain- and loss-expectation, anchor-bias or feedback mechanisms and learning functions must be taken into account. The talk gives an overview of the phenomena that are relevant in shared decision-making with AI, e.g. in the fields transport and logistics or finance.