Could artificial intelligence improve astronaut debriefing processes? A review from psychological and human–computer interaction perspectives

Debriefing is a cornerstone of astronaut training and performance, functioning as a structured reflective practice that transforms operational experience into learning. In the extreme environments of human spaceflight, effective debriefing supports not only safety and performance improvement but also psychological adaptation under high cognitive and physical demands. However, traditional debriefing practices rely on human facilitation, subjective recall, and fragmented data sources, which can limit their consistency and analytical depth. Recent developments in artificial intelligence (AI), like large language models, provide new opportunities to enhance debriefing processes. These technologies could support automated transcription, data integration, structured feedback generation, and self-regulated learning. Drawing on ecological psychology and self-regulated learning theories, this perspective paper examines how AI may support debriefing in astronauts training, as well as the potential psychological implications of AI applications. We argue that the integration of AI may enable such precise and data-informed feedback, which could help to individualize the debriefing process. However, the integration of AI may also raise important challenges related to data privacy, human oversight, technical implementation, and ethical use in such contexts.