Every day, thousands of images and signals are collected at sea. Sonar, buoys, satellites, and cameras installed on ships generate enormous amounts of data. Artificial intelligence (AI) is already being used to interpret this information. For example, to detect the presence of dolphins in real time to prevent bycatch, to estimate biodiversity indicators, or to automatically identify species caught onboard fishing vessels and improve fisheries management models. But behind this technological transformation emerges a key question: can we fully trust what AI says when the health of the ocean is at stake?
