In the CONFIDENCE project, we developed an agent based model (ABM) to simulate the decision making process involving stakeholders of different interests. Our model aims to support decisions on the most suitable protection strategies in different accident phases. The intelligent agents and the models of the negotiation/voting process are described in the paper. Given five scenarios, the numerical results from the computational implementation of the ABM are visualized and analysed in order to better understand the negotiation and voting processes. Our ABM can be expanded in order to support the decision making processes of many different stakeholders of various types of risk management apart from nuclear and radiological emergency management.