Artificial Intelligence in Safety-Critical Aerospace Systems: Applications, Challenges, and Implications

Abstract This paper investigates the role of artificial intelligence (AI) in safety-critical aerospace systems. It emphasizes its function not merely as an automatic aid but as an increasingly autonomous decision-making agent embedded within complex engineering environments. The investigation of AI applications is centered around Reinforcement Learning techniques and other approaches that are considered supporting this concept. An interdisciplinary framework is adopted, integrating technical analysis of AI applications with system-level considerations in design and operations. In order to provide a broad picture of the implications of the use of AI-based tools into aerospace systems, and as a novelty in the field, it includes issues of governance, legal responsibility, accountability, and ethics. The study is grounded in three application domains —hypersonic atmospheric reentry, autonomous satellite systems, and nuclear space propulsion— selected for their high-risk operational contexts and minimal tolerance for failure. Across these domains, the paper examines how AI-driven decisions directly influence system performance in all phases of a system’s life, from design through deployment and operation. It highlights the need for robust frameworks to ensure reliability, transparency, and responsible deployment. A set of thematic constructs are developed which demonstrates how they manifest across the three application domains. The aim of this paper is to inform current and future system designs and policy frameworks, through an interdisciplinary approach grounded in current practices.