“…To the best of our knowledge, no work targeting the autonomous evaluation of risk indicators exists yet. In this section, we will, therefore, discuss existing work for the individual parts of FedCRI, covering the detection of risks on smartphones [29], [53], [23], [32], [44], [26], [56], coordination of different clients for threat intelligence [28], [9], [17], [13] and leveraging ML for detecting risks and anomalies [49], [42].…”