Interoperability is broad and complex subject being the most critical issue facing businesses that need to access information from multiple systems. When talking about achieving interoperability with other systems it is not uncommon to neglect the concept of unwanted interoperability. Such interoperability can result in fault decision making based on counterfeit data produced by hostile system that is interoperable with the system observed. Nowdays information systems are often created by integrating existing subsystem. Integration becomes more important than development due to short time in disposal between the contract signature and on site implementation. Developed information system is composed of elements interacting and exhibiting emergent properties. Complexity science is good candidate to describe such a system. It is crucial to be able to model the system before deciding on element to be used in the integration or development process. Toll collection system is complex system incorporating business information and equipment control system characteristics. Typical example of unwanted interoperability in toll collection system is payment authorisation fraud. Research in this paper is based on highway toll collection system analysis as representative of hierarchical heterogeneous systems. A mechanism to detect unwanted interoperability using information fusion collected from different information system levels is presented.