Norms of negative consequences might be practiced in a society for a long time. These norms are mostly related to strong beliefs that prevent individuals from violating or abandoning them. In this case, these norms must be removed to reach a stable and benevolent society. Negative norms removal is presented in social science as a part of society evolution, yet it is not modeled in normative multi-agent systems. Norm removal in social science is done in two main stages. The first stage is called collective belief change. The second stage is called collective action. In this paper, we model these two stages into five processes which are, norm negativity realization, collective belief change, norm removal decision, choosing removal monitoring authority, and removal process. When the five stages are completed, agents make their own decision either to delete or not to delete the norm from their cognitive structure depending on their internal system status.