Workshop on Socially Aware and Cooperative Intelligent Systems

The use of conversational AI, such as ChatGPT or DeepSeek, is rising rapidly across many personal and professional settings. These AI models not only enable natural-language interactions but are also increasingly equipped with realistic synthetic voices and other human-like cues that mimic interpersonal communication. Research, including recent studies from Prof. Mara’s Robopsychology Lab, shows that many people readily anthropomorphize AI language models, attributing emotionality, personality, and intentionality to them. While anthropomorphism can help reduce complexity and satisfy the human need for social relatedness, it may also raise concerns. Studies suggest that individuals who more strongly anthropomorphize AI are more likely to endorse granting such systems rights and to judge AI-generated recommendations as more competent and trustworthy, thereby increasing the risk of over-reliance on potentially flawed outputs. Although anthropomorphism of non-human entities is very widespread, it varies with individual and contextual factors. Findings from the Robopsychology Lab indicate that people who feel lonely or have limited AI literacy can be especially prone to anthropomorphism. This talk will examine the psychological mechanisms underlying AI anthropomorphism, its implications for both human-AI and human-human interaction, and practical strategies for de-anthropomorphizing and improving AI literacy in an increasingly AI-integrated world.

Martina Mara is a Full Professor of Psychology of Artificial Intelligence & Robotics and Head of the Robopsychology Lab at Johannes Kepler University Linz, Austria. She earned her PhD in Psychology from the University of Koblenz-Landau (2014), focusing on user acceptance of highly human-like robots, and received her habilitation (venia docendi) from the University of Nuremberg (2022). Trained as an empirical psychologist with an interdisciplinary outlook, her research examines how people perceive, trust, and collaborate with intelligent systems—and how system design can foster human autonomy, creativity, and well-being. Before joining JKU in 2018, Martina spent many years outside academia, working in media companies and a design studio, and later at the Ars Electronica Futurelab, where she collaborated with international partners, including major companies across Europe and Japan, at the intersection of technology, research, and the arts. She is a co-founder of the Initiative Digitalisierung Chancengerecht (IDC), serves on the board of the Austrian Research Promotion Agency (FFG), and has received several honors, including the Vienna Women’s Prize, a Futurezone Award, and the Kaethe Leichter Prize for outstanding contributions to women’s and gender research from the Austrian Federal Ministry for Digital and Economic Affairs. Alongside her teaching and research, she is passionate about science communication, giving public talks, writing commentary, and creating interactive exhibitions on her lab’s topics.

Human-Agent Interaction (HAI) has traditionally centered on dialogue between humans and artificial agents in relatively controlled settings. In this talk, I will present a broader perspective on Ecological AI—agents that adapt not only to humans, but also to the diverse environments in which interaction occurs. The journey begins with robot audition, enabling robots to localize, separate, and understand speech in noisy real-world acoustic scenes. We then extend auditory intelligence to drone audition for disaster response, where agents must support human rescuers under extreme uncertainty. Beyond audition, I will highlight efforts on two-way sign language translation and generation, enabling inclusive communication that bridges spoken and signed languages, and allowing agents to interact naturally with people who rely on sign. Finally, I will turn to wildlife and ecosystem monitoring, where agents listen to and interpret animal vocalizations, connecting humans with the voices of nature. Together, these studies illustrate how HAI can evolve into human–agent–environment interaction, expanding from human-centered conversation to inclusive, resilient, and ecological forms of collaboration.

Kazuhiro Nakadai received his B.E., M.E., and Ph.D. degrees in electrical and information engineering from the University of Tokyo in 1993, 1995, and 2003, respectively. He worked at Nippon Telegraph and Telephone (1995–1999), the Kitano Symbiotic Systems Project, ERATO, JST (1999–2003), and Honda Research Institute Japan as a principal scientist (2003–2022). He is currently a professor at the Department of Systems and Control Engineering, Institute of Science Tokyo (formerly Tokyo Institute of Technology). He also held visiting and special appointments at Tokyo Institute of Technology (2006–2022) and served as a guest professor at Waseda University (2011–2018). His research interests include artificial intelligence, robotics, signal processing, computational auditory scene analysis, multimodal integration, and robot audition. He served on the executive boards of the Japanese Society for Artificial Intelligence (2015–2016, 2024–2025) and the Robotics Society of Japan (2017–2018). He is a Fellow of IEEE and RSJ.

In recent years, AI companions have gained remarkable popularity worldwide. What began as relatively simple chatbots or entertainment applications has rapidly developed into systems that many people now turn to for emotional support, daily conversation, and even romantic companionship. For some individuals, these AI partners are not only a source of comfort but also function as meaningful figures in their personal lives. This development highlights a significant cultural and technological shift: artificial agents are no longer perceived solely as tools, but increasingly as “partners” in human relationships. The growing prevalence of AI companions prompts us to reconsider what it means to form emotional attachments, and how such relationships might influence human-to-human connections in broader society. This workshop presentation will examine these questions by exploring recent findings on the psychology of virtual intimacy of human and virtual agents. These findings highlight the psychological potential of virtual romance and the importance of understanding how emerging technologies are reshaping human relationships.

Mayu Koike is an assistant professor at the Institute of Science Tokyo. She obtained her PhD in Psychology from The University of Edinburgh. Her work focuses on the relationships between people and virtual agents. She is particularly interested in how people form strong attachments (love and romantic relationships) with virtual agents, and the potential for improving psychological well-being within this context.