Because of the complexity of real-world problems, large-scale group decision making has become a research topic of great interest in the field of decision science. Differences of opinion in a large group are highly likely. Sometimes, decision makers are unwilling to adjust their opinions to promote consensus. It is hence necessary to establish a consensus model for the effective management of opinion differences and non-cooperative behaviors. More importantly, the credibility of the adjustment information must be ensured. In this paper, we present a confidence consensus-based model for large-scale group decision making that provides a novel approach to addressing non-cooperative behaviors. First, some new concepts are proposed, including the collective adjustment suggestion and rationality degree. Then, we combine the rationality and non-cooperation of the adjustment information to construct the concept of a confidence level. This confidence level measures the impartiality and objectivity of the adjustment information and is the basis for managing non-cooperative behaviors. We then establish a mechanism for addressing non-cooperative behaviors. Finally, we present a case study that illustrates that the proposed model is feasible and effective. A comparative analysis reveals the features and advantages of this model for managing large-scale group decision making.