The demands of transportation by way of wheeled vehicles on roadways is increasing as society becomes increasingly digital. More people and products become connected and obligated, then subsequently must be moved to meet those obligations. As demand to move those people and goods outpaces reasonable prices, automation becomes the apparent solution. First and foremost is the concern of safety for the people in and around the automated vehicles. The first ring in safe operation of autonomous vehicles is decision making from the presented data, usually represented by a machine learning model. Even if the decision-making process of the car is impossibly perfect given input data, the data itself comes from the environment, which means it is an external attack vector. Such attacks are made on input data to the autonomous vehicle’s sensors and must be thoroughly examined and cleaned of malicious manipulation in order to ensure the safe operation of the autonomous vehicle in public spaces. This can be done through a panel of machine learning models to label data on specific attributes, flagging malicious data. A second solution is adding multiple sensors from different angles and depths to verify the veracity of incoming signals and observations. A third solution is to build infrastructure into the roadways to act as supplemental sensors to make observations on behalf of the car. A fourth solution is to allow inter-car communication as a form of environmental input verification.