The critical nature of vehicle steering safety underscores the importance of exploring potential cyber-physical attacks on modern steering systems. This chapter delves into how such attacks can destabilize the interactive dynamics between human drivers and vehicle haptic shared control (HSC) steering systems. Contrary to traditional research that focuses on stabilizing these dynamics, this study explores their vulnerability. It illustrates the destructive potential of cyber-physical attacks on HSC systems by showing how an attacker could employ a delay-dependent feedback control law to disrupt them. By introducing delayed reference angular commands to the driver input device and the vehicle’s steering control assembly, the attacker effectively destabilizes the HSC system. Additionally, this chapter discusses how machine learning algorithms could enhance the resilience of HSC steering systems by detection and mitigation of such attacks.