Feedback tampering is a critical cyberattack targeting the monitoring functions of instrumentation and control (I&C) systems. By manipulating feedback data from field equipment, these attacks can either mask the reporting of real anomalies or generate false alarms to distract the operators, rendering the I&C function unreliable, as stated in IAEA NSS 33-T. To mitigate these risks, this study proposes a heterogeneous sensing approach. Employing multiple sensor types to monitor diverse but correlated attributes of the same equipment or process significantly increases the difficulty of executing coordinated tampering across all sensing channels. Heterogeneous sensing leverages the inherent correlation between these data streams to detect anomalies effectively. A test bed was designed to evaluate the approach, involving a motor controlled by a variable frequency drive (VFD) with simulated tampering. Heterogeneous sensors, including three thermometers and a vibration sensor, capture complementary aspects of the motor’s operation. The data correlation from these sensors aids in detecting feedback manipulation. A deep learning tamper detection algorithm is developed to identify discrepancies caused by malicious feedback manipulation. This study demonstrates how heterogeneous sensing, coupled with advanced AI techniques, can enhance the cybersecurity of I&C systems.