The rapid digitization of the automotive industry has ushered in advanced connectivity features, autonomous functionalities, and data-driven decision-making processes. However, this technological transformation has also introduced significant cybersecurity vulnerabilities within automotive embedded systems, where traditional security frameworks struggle to address evolving threats. Cyberattacks targeting vehicle control units, in-vehicle networks, and external communication interfaces can compromise passenger safety, disrupt mobility services, and undermine consumer trust. A broader understanding of these risks highlights the urgent need for robust defense mechanisms capable of adapting to sophisticated intrusion attempts. Artificial intelligence (AI)-driven intrusion detection systems (IDS) present a promising approach, leveraging machine learning and deep learning techniques to identify anomalous patterns in vehicular data streams in real time. These systems not only enhance detection accuracy but also enable adaptive responses to previously unseen attack vectors. Narrowing the focus further, secure communication protocols, including cryptographic authentication and lightweight encryption schemes, play a pivotal role in safeguarding data exchange across in-vehicle networks and vehicle-to-everything (V2X) infrastructures. Integrating AI-driven IDS with resilient communication frameworks offers a layered security model that strengthens embedded system resilience against cyber threats while ensuring compliance with stringent automotive safety standards such as ISO/SAE 21434. This paper emphasizes the synergy of AI-based threat detection with secure communication architectures as a comprehensive strategy to mitigate cybersecurity risks. By aligning technological innovations with regulatory and industry practices, the automotive sector can address emerging challenges proactively, ensuring the safe deployment of connected and autonomous vehicles in an increasingly hostile cyber environment.