This study x-rays safety and decision-making algorithms in AI-driven autonomous vehicles. The objectives of the study is to contribute to the ongoing dialogue on safety and decision-making considerations of AI-driven autonomous vehicles, scrutinize existing literature related to the ethical considerations of AI-driven autonomous vehicles to identify key issues and potential solutions, and determine if autonomous-vehicles are best for the roads of the world. This study employs the case study approach in its research design. The study analyzes case studies of real-time deployments of AI-Driven Autonomous vehicles, examines how safety and decision-making concerns are handled in practice. The findings from the literature review, regulatory analysis, case studies, and stakeholder interviews were analyzed to identify key areas, emerging trends, and areas of divergence or consensus. Our findings throw light on the different views regarding decision-making reliability, a key component in establishing confidence. Engineers typically place considerable confidence in a system’s decision-making patterns following rigorous testing and a proven success track record. On the other hand, designers often put transparency and widespread adoption first, which may not correlate with decision-making reliability directly. While safety is seen as a universally prioritized aspect by both engineers and designers, it alone is not sufficient enough to institute complete confidence, therefore, the incorporation of safety and decision-making reliability is necessary to build user confidence in Artificial Intelligence systems for autonomous vehicles. The study recommends that there should be enhancement of adversarial robustness in artificial intelligence algorithms and the implementation of cybersecurity protocols. Improved sensor technologies play a crucial role in improving perception capability, and human-AI relation design becomes paramount for driving trust and understanding amongst users.

Algorithms, Artificial Intelligence, considerations, decision-making, ethics, reliability, safety, self-driving vehicles.