On board vessels, the Automatic Identification System sends and receives localization messages and enables vessels to better understand their surroundings. Initiated by the Safety Of Life At Sea convention, this system is used for navigation security and safety, boarding prevention, fleet control or traffic control. Some of the messages broadcasted contain errors, falsification and undergo spoofing that weaken the capaci-ties of the system to achieve its goals. This paper presents a risk analysis study of the Automatic Identification System that leads to the identification of circa 350 threat scenarios. A typology of anomalies is proposed, alongside with a methodology for anomaly detection. The objectives are the determination of the false mes-sages and the improvement of both the effectiveness of the system as a security system and the maritime situa-tional awareness.