This volume contains papers presented at the 2nd International Conference on Modeling Decisions for Arti?cial Intelligence (MDAI 2005), held in Tsukuba, Japan, July 25-27. This conference follows MDAI 2004 (held in Barcelona, Catalonia, Spain), the proceedings of which were also published in the LNAI series (Vol. 3131). The aim of this conference was to provide a forum for researchers to discuss about theory and tools for modeling decisions, as well as applications that - compass decision-making processes and information fusion techniques. In this second edition, special focus was given to applications related to risk, security and safety. The organizers received 118 papers, from 14 di?erent countries, 40 of which are published in this volume. Each submission received at least two reviews from the Program Committee and a few external reviewers. We would like to express our gratitude to them for their work. The plenary talks presented at the conference are also included in this volume. The conference was supported by the Department of Risk Engineering of the University of Tsukuba, the Japan Society for Fuzzy Theory and Intelligent Informatics (SOFT), the Catalan Association for Arti?cial Intelligence (ACIA), the European Society for Fuzzy Logic and Technology (EUSFLAT) and the Generalitat de Catalunya (AGAUR 2004XT 0004).