The development of autonomous multi-unmanned aerial vehicles, or multi-UAVs, represents significant progress in various fields, including the military, business, and civilian sectors. As this innovative field develops, the necessity of creating models that guarantee the fundamental components of security, privacy, dependability, effectiveness, and safety is growing. Based on this observation, this research suggests a reputation-based team formation (RTF) approach that rates each UAV according to the level of services it provides. Such a system is essential for critical multi-UAV Beyond Visual Line of Sight (BVLOS) operations requiring high dependability, effectiveness, safety, and privacy protection. To validate RTF’s effectiveness, we focus on a simulation scenario where multi-UAVs are deployed to localize a target in a forest where a fire is spreading. The criticality of this operation is underscored by the need to locate the target before the fire reaches it. Our results demonstrate that RTF successfully enables cooperative UAVs to block malicious ones and significantly reduce their influence, even in the presence of multiple compromised UAVs.