The rise of autonomous vehicles is transforming transportation. However, the susceptibility to cyberattacks poses a significant problem. Cyberattacks can exploit weaknesses in sensors and control systems. This may instigate accidents and compromise passenger safety. Recent studies have shown a 30% increase in cyberattacks on self-driving cars, with hackers targeting GPS signals and sensor data to manipulate vehicle behavior. To address this issue, we came up with a robust cybersecurity system. It will detect and respond to threats in real time. Our solution includes two main layers. The first layer secures the vehicle’s control systems and communication channels from any unauthorized access. We can ensure that the data which is being transmitted between the vehicle and external systems remains confidential and untampered, by establishing secure communication protocols. The second layer uses a machine learning model to monitor sensor data for anomalies. This model is trained on large amount of dataset to learn normal or regular patterns and quickly identify any kind of deviations that may indicate a cyber threat. It allows us to respond to threats in milliseconds, keeping the vehicle safe and secure even under attack. In simulations, the proposed system achieved a 93.3% accuracy in detecting and mitigating adversarial manipulations. Moreover, real-world testing demonstrated a 25% reduction in cyber-related malfunctions, showcasing the system can effectively protect vehicles from being hacked or disrupted. Our research highlights how important it is to have real-time threat detection and machine learning to strengthen the cybersecurity framework of autonomous vehicles.