In recent years, assembly line design methodologies have included the possibility of sharing tasks between stations. These approaches increase the throughput, improving the workload balance among the stations. Sharing operations can support the management of assembly lines with task redundancy. The redundancy of the tasks between two consecutive stations improves the robustness of the assembly line to unforeseen events. A control policy that allocates the shared tasks between two adjacent stations is crucial for the performance of the assembly line. This paper proposes a fuzzy approach to distributing the workload of the shared tasks between two consecutive stations. The fuzzy model allows for the distribution of the workload between the stations in a more gradual way. Simulations were conducted to allow for a performance evaluation with a model proposed in the literature. The results show how the fuzzy model improves all the performance measures evaluated; in particular, when the inter-arrival of the raw items is rather stable. In these conditions, the reduction of the average time and its standard deviation is about 50% to 80%, while the reduction of the work in process is about 60% to 90%. These benefits are obtained with a relevant increase of the variations of the average utilization of the stations.