A novel approach to the important problem of aggregation of individual fuzzy preferences of a group of agents in a group decision process is presented. It is different than the traditional, widely employed aggregation based on the averaging, the weighed conjunction/disjunction, etc. of the respective preference degrees between pairs of options which stand for the uniformity of the group of agents in the sense that all testimonies of the agents are to be accounted for, even if the agents (and maybe also options) are assigned various importance degrees. The novel approach proposed here assumes that the group is not uniform in the above sense which implies that the testimonies of the most important agents are crucial and decisive, so that they should be followed, possibly ignoring testimonies of lower level agents, as it often happens in business or the military. Yager’s approach to the so-called prioritized aggregation is employed. The results are very promising and can yield new vistas and perspectives for (fuzzy) preference based group decision making in complex groups of agents, notably authoritarian hierarchies.