YEREVAN, Armenia — On September 3, the American University of Armenia hosted an event exploring how generative artificial intelligence (AI) models approach one of the most difficult questions surrounding contemporary conflict: What might constitute a just end to the Russo-Ukrainian War? The event was organized by the Turpanjian Institute of Social Sciences and the Master of Arts in International Relations and Diplomacy program.
The presentation was delivered by Dr. Arseniy Kumankov, a researcher specializing in war studies and the ethics of war and peace. He is currently a research fellow at the MALVA Center for the study of war, language, and memory at the University of Mainz, Germany.
The presentation examined how large language models (LLMs) reason about justice, victory, and peace when confronted with the moral and political dilemmas surrounding the conclusion of war.
As generative AI increasingly becomes a medium through which people encounter political information, interpret international events, and form opinions about war and peace, questions about how these systems process complex political and ethical issues are becoming increasingly important. The presentation considered whether AI models can offer consistent approaches to such questions or whether different models produce fundamentally different visions of a just peace.
During the presentation, identical dilemmas concerning the end of the Russo-Ukrainian War were presented to several leading AI models, including ChatGPT, Claude, Gemini, and DeepSeek. By comparing their responses, the discussion explored the extent to which the models converge on similar understandings of justice and peace and where their reasoning diverges.
The analysis highlighted the ways in which LLMs draw upon and recombine existing human political narratives and ethical frameworks. Trained on vast bodies of human-generated text, AI models can reproduce established ways of understanding war and peace while also generating unexpected combinations of existing arguments.
A central question of the discussion was therefore not only what answers AI models provide, but also which political narratives remain visible through algorithmic processing and which are marginalized or silenced. Examining these patterns can offer insight into the assumptions and frameworks embedded in the information environments increasingly shaping public understanding of international conflicts.
The event provided an opportunity for participants to consider the growing role of AI in interpreting political and ethical questions and to reflect on how algorithmic systems may influence contemporary understandings of war, justice, and peace.
Founded in 1991, the American University of Armenia (AUA) is a private, independent university located in Yerevan, Armenia, affiliated with the University of California, and accredited by the WASC Senior College and University Commission (WSCUC) in the United States. AUA provides local and international students with Western-style education through top-quality undergraduate and graduate degree and certificate programs, promotes research and innovation, encourages civic engagement and community service, and fosters democratic values.
