The recent advances in Artificial Intelligence (AI), among many other use cases, allows consumers to voice their concerns more easily. For instance, AI is used extensively for identifying infringements to consumer rights, preparing complaints and even submitting them to the authorities. However, these authorities are stripped from the ability to use similar Machine Learning-based tools due to concerns about correctness. Traditionally, one resorts to expert systems for providing robust and correct solutions. Such systems are not as wide-spread as desirable though. One key issue is that such systems are very powerful, making it often difficult for their users, i.e., jurists, to fully understand them without significant effort. Here, we consider one specific use case, consumer complaints, and propose an approach that puts intelligibility and trust first, which allows a jurist to partially automate the assessment of consumer complaints.