In-vehicle Electronic Control Units (ECUs) rely on continuous communication over in-vehicle networks, such as the Controller Area Network (CAN). As the demand for connectivity increases, so does the complexity of these networks. In recent years, vehicle security has become a critical focus for the industry, driven by incidents involving malicious attacks. The AUTOSAR software development framework addresses some security concerns through mechanisms like Freshness Values (FV) in Secure Onboard Communication (SecOC). However, these measures remain vulnerable to sophisticated attacks such as replay and masquerade attacks, which adversaries can exploit to compromise system integrity and extract sensitive information. To address these vulnerabilities, we propose the Message Prediction Table (MPT), a novel approach designed to enhance communication security among ECUs in the CAN network. The MPT predicts the next expected message each ECU will receive, bolstering AUTOSAR’s defenses against replay and masquerade attacks. Our experiments evaluate the MPT’s effectiveness about security, its impact on communication overhead, and its compatibility with AUTOSAR standards. Results indicate that while the MPT introduces a modest increase in communication overhead, it significantly enhances the network’s resilience against adversarial attacks, offering a robust solution for securing in-vehicle communication.