Self-driving cars are gaining momentum despite the number of considerable technical and human factor issues that remain controversial. Human acceptance and trust of an automated vehicle to transport people in traffic environments under different driving conditions is a challenging task. Perceived, as well as actual safety will play a major role in accepting automated vehicles. Perceptions however vary from one individual to another due to different reaction times, speed perception, and time constants during dynamical changes etc. The closer the automated vehicle dynamics are with those of a manually driven vehicle the more likely that the comfort level of the automated vehicle user will improve. In this paper we review these issues and discuss how the autopilot personalization feature can help to improve both the perceived and actual driving safety and comfort. We present methodology that allows automatic autopilot personalization based on driver performance models. The methodology takes into account driver's preferences for a particular trip and manual parameters fine-tuning by the driver. We demonstrate how the methodology can be applied on an example of adaptive cruise control and automatic lane change personalization. We support the example with data collected on an experimental vehicle.