Areas of Interest

We welcome the submission of technical papers describing significant and original research on all aspects of the theory and practice of autonomous agents and multiagent systems. If you are new to this community, then we encourage you to consult the proceedings of previous editions of the conference to fully appreciate the scope of AAMAS. At the time of submission, you will be asked to associate your paper with one of the following areas of interest:

    Learning and Adaptation (LEARN)
    Generative and Agentic AI (GAAI)
    Game Theory and Economic Paradigms (GTEP)
    Coordination, Organizations, Institutions, Norms, and Ethics (COINE)
    Search, Optimization, Planning, and Scheduling (SOPS)
    Representation, and Reasoning (RR)
    Engineering and Analysis of Multiagent Systems (EMAS)
    Modeling and Simulation of Societies (SIM)
    Human-Agent Interaction (HAI)
    Robotics and Control (ROBOT)
    Innovative Applications (IA)

Additionally, AAMAS 2026 includes several special tracks.  You can find more information about these tracks under the Calls menu item.

Learning and Adaptation (LEARN)

Area Chairs: Bo An, Vincent Corruble, Yali Du, Ferdinando Fioretto,  Sarah Keren, Yasser Mohammad, Francisco Cruz Naranjo, Tan Minh Nguyen, Karthik Sankararaman, and Kun Shao

Topics:

    Reasoning and learning under uncertainty
    Supervised learning
    Unsupervised and representation learning
    Reinforcement learning
    Multiagent learning
    Evolutionary and biologically inspired algorithms
    Learning agent capabilities
    Learning agent-to-agent interactions
    Human-in-the-loop learning
    Few-shot learning
    Distributionally-robust learning
    Adversarial learning

Description:

Papers in this area focus on aspects of single agent and multiagent learning and communication. This includes all areas of machine learning such as unsupervised, supervised, and reinforcement learning and methods such as evolutionary and adversarial approaches. Note that there is also an area on Generative Agents and Agentic AI which is described below.
Generative and Agentic AI (GAAI)

Area Chairs: Prashant Doshi, Kate Larson, and Karl Tuyls

Topics:

    Agency and learning in LLMs
    Learning for value alignment
    Reinforcement learning from human feedback (RLHF)
    Planning or learning for agentic processes/workflows
    Agentic decision aids
    Verification and safety of LLMs/agentic systems
    Modeling and analysis of generative AI agents
    Multi-agent training of LLM agents
    (Multi-turn) Instruction following of agents
    Evaluation of LLMs/agentic systems
    Cooperative and coordination of generative agents

Description:

This area focuses on advancing agentic AI, exploring how generative agent models learn, align with human values, and interact effectively, both with humans and with other agents. Key themes include enabling agency in LLMs and other generative models through methods like reinforcement learning from human feedback, designing agents for complex workflows, and decision-making while emphasizing critical aspects like verification, safety, multi-agent training, agent-to-agent communication and evaluation of these systems.
Game Theory and Economic Paradigms (GTEP)

Area Chairs: Niclas Boehmer, Noam Hazon, Omer Lev, Miming Li, Shuai Li, Neeldhara Misra, Svetlana Obraztsova, Maria Silvia Pini, and Alan Tsang

Topics:

    Auctions and Mechanism Design
    Bargaining and Negotiation
    Behavioural Game Theory
    Evolutionary Game Theory
    Non-Cooperative Games: Equilibrium Concepts
    Non-Cooperative Games: Computational Issues
    Non-Cooperative Games: Theory and Applications
    Voting and Preference Aggregation
    Social Choice and Social Networks
    Judgment Aggregation
    Fair Allocation
    Tournaments
    Matching
    Digital and Liquid Democracy
    Coalition Formation
    Cooperative Games

Description:

This area encompasses research on cooperative and non-cooperative games, social choice, and mechanism design. It is particularly interested in various computational aspects such as novel algorithms, equilibrium computation, and efficient outcomes. The area welcomes both theoretical explorations and analysis related to game theory, mechanism and market design, and social choice; as well as showcasing practical applications of these topics.
Coordination, Organizations, Institutions, Norms, and Ethics (COINE)

Area Chairs: Nirav Ajmeri, Reyhan Aydoğan, and Jaime Simão Sichman

Topics:

    Coordination and teamwork
    Social network analysis
    Norms and normative systems
    Organizations and institutions
    Non-strategic coalition or team formation
    Communication, including using natural language
    Policy, regulation, and accountability
    Safety, robustness, trust, and reputation
    Ethical considerations, including bias, equity, fairness, privacy, safety, security, transparency
    Values and preferences
    Agreement Technologies: Negotiation and Argumentation
    Responsible socio-technical systems
    Explainability and interpretability of norms and ethics in human-agent teams
    Ethical challenges of using LLMs for coordination and cooperation

Description:

Research in agent and multiagent systems has a long history of developing techniques that balance agent autonomy, adaptation, and distributed social reasoning with system-level considerations such as organizational and institutional policy enforcement addressing safety, security and fairness considerations. Teamwork and human-machine cooperation has an increased relevance with the transformation of our societies into socio-technical systems. We need to ensure transparency, foster trust, and align social reasoning with societal norms and expectations. We also need to ensure human-machine and machine-machine cooperation is fostered responsibly, within an adequate accountability system and in alignment with the ethical values of individuals concerned. We encourage the submission of papers that highlight the design, development, evaluation, simulation, and analysis of novel, innovative, and impactful research on issues related to the above topics.
Search, Optimization, Planning, and Scheduling (SOPS)

Area Chairs: Roman Barták, Sara Bernardini, and Filippo Bistaffa

Topics:

    Single-agent planning and scheduling
    Multi-agent planning and scheduling
    Decentralized planning and scheduling
    Planning under uncertainty
    Combinatorial optimization
    Constraint programming
    Distributed constraint reasoning
    Resource and task allocation
    Non-strategic coalition formation
    Plan and goal recognition
    Human-aware planning and scheduling
    Knowledge Representation/Engineering for SOPS

Description:

This area includes theoretical or experimental contributions to search, optimization, planning, and scheduling in single- and multi-agent systems. Important subfields include decentralized planning, planning under uncertainty, combinatorial optimization, distributed constraint reasoning, resource and task allocation, and non-strategic coalition formation.  Approaches for planning and scheduling that explore the use of machine learning and/or foundation models in conjunction with classical techniques are encouraged. Likewise, all approaches to single- and multi-agent planning, including motion and path planning, and their interplay with other agent components, are relevant.
Representation and Reasoning (RR)

Area Chairs: Roberta Calegari, Brian Logan, Vanina Martinez and Valentina Tamma

Topics:

    Neurosymbolic approaches
    Argumentation
    Agent theories and models
    Explainability
    Logics for agent reasoning
    Ontologies for agents
    Reasoning about knowledge, beliefs, goals, actions, plans, and change in multiagent systems
    Reasoning and problem solving in agent-based systems
    Verification of agents and multiagent systems

Description:

This area includes theoretical or experimental contributions to knowledge representation and reasoning in single-agent and multi-agent systems. Knowledge representation is to be interpreted broadly, encompassing formal approaches—such as epistemic, strategic, and description logics—as well as data-driven techniques like representation learning. We also welcome contributions involving semantic frameworks such as knowledge graphs and ontologies, particularly where they support agent interoperability, communication, or reasoning. Representation and reasoning in complex settings often require the integration of perception and sensing, enabling agents to operate effectively in dynamic and uncertain environments. Relevant reasoning paradigms include automated reasoning, theorem proving, verification-based approaches, probabilistic inference, and neurosymbolic methods, provided they are applied to or inspired by reasoning in agent and multi-agent systems. Special emphasis is placed on methods that promote trustworthiness and responsible AI. This includes techniques for managing uncertainty and bias in representations, ensuring accountability and transparency, and embedding fairness and explainability into decision-making processes.
Engineering and Analysis of Multiagent Systems (EMAS)

Area Chairs: Leandro Buss Becker, Angelo Ferrando and Zahia Guessoum

Topics:

    Requirements capture & formal specification of multi-agent systems
    Programming paradigms & languages for autonomous agents
    Runtime infrastructures and deployment platforms for scalable MAS (cloud, edge, and hybrid)
    Continuous verification, validation & certification pipelines
    Testing, debugging and DevOps for large-scale agent ensembles
    Scalability, fault-tolerance & performance engineering of MAS platforms
    Engineering self-adaptive agents: lifecycle management, continuous evolution, and deployment pipelines
    Interoperability, business agreements & agent-to-agent protocols
    Declarative, logic-based and BDI agent programming & architectures
    Engineering MAS-based simulations for rigorous analysis & experimentation
    Sociotechnical governance tools for norms, ethics & accountability
    Human-centred engineering for usability, transparency & explainability
    Open-source toolchains, benchmarks & reproducible MAS testbeds
    Engineering learning agents: platform design, online adaptation
    Engineering MAS with LLM methods
    Hybrid symbolic-subsymbolic agent systems: engineering architectures and platforms for neurosymbolic reasoning
    Benchmarks, evaluation methodologies, and reproducible workflows for MAS
    Data-driven engineering processes for design, testing, and adaptation of agent systems
    Engineering MAS-based autonomous systems

Description:

This area invites contributions that advance the engineering of agents and multi-agent systems through general-purpose software abstractions, programming paradigms, methodologies, and system infrastructures. We welcome work spanning the full software engineering lifecycle—from requirements capture and formal specification to testing, validation, deployment, and evolution. Emphasis is placed on engineering-oriented approaches that support the development of robust, scalable, and maintainable agent-based systems, including those integrating symbolic reasoning and learning capabilities (e.g., neurosymbolic agents, reinforcement learning agents, or LLM-based agents), provided the focus remains on engineering challenges and solutions. Contributions that demonstrate the impact of these approaches in concrete application domains or in synergy with emerging technologies are also encouraged.
Modeling and Simulation of (Artificial) Societies (SIM)

Area Chairs: Shah Jamal Alam, Franziska Klügl and Fabian Lorig

Topics:

    Modeling for agent-based simulation
    Simulation of complex systems
    Modeling of societies and social simulations
    Public policies and evaluation
    Simulation techniques, tools and platforms
    Analysis of agent-based simulations
    Calibration, Verification and Validation of agent-based simulation systems
    Robustness, Reliability and Trustworthiness of agent-based simulations
    Design and implementation of large-scale simulation
    High-performance computing and frameworks
    Interactive and participatory simulation
    LLM-based simulation

Description:

Artificial societies are computer simulations or models that are created to emulate and research the behavior of intricate social systems. These societies simulate the interactions and dynamics of people, animals or other entities to understand how individual behaviors lead to emergent structures and interactions. Agent-based models of artificial society provide a way to analyze the impact of regulations, incentives and other interventions that help to understand the complex dynamics of society as a whole. The area aims to find efficient solutions to model and simulate complex societal systems using agents-based models. Important application areas include ecology, biology, economics, transportation, management, organizational, and social sciences in general. In these areas, agent theories, metaphors, models, analysis, experimental designs, empirical studies, and methodological principles, all converge into simulation as a way of achieving explanations and predictions, exploration and testing of hypotheses, and better system designs.
Human-Agent Interaction (HAI)

Area Chairs: Beatrice Biancardi, Brittany Duncan, and Matthias Scheutz

Topics:

    Human-agent interaction
    Agent-based analysis of human interactions
    Socially interactive agents
    Trust and explainability in human-agent interactions
    Human-robot interaction and collaboration
    Social robotics and social interactions
    Mixed-initiative and shared autonomy in human-agent interactions
    Groups of humans and agents
    Agent models and architectures for interaction with humans
    Design for human-agent interaction
    Virtual humans

Description:

Human interaction with artificially intelligent agents is becoming more commonplace, as such, developing and evaluating agents that can understand humans’ dynamics to support competent interaction. Significant challenges arise when transitioning from pure multiagent systems to hybrid systems that need to incorporate bi-directional human and agent interactions and sustain different competitive or collaborative situations. Agents need new models and architectures to better address the interaction with humans, including perception and recognition of humans’ internal states and activities at different levels, interaction modalities that support true coordination, and incorporation of human factors and ethics concerns. The design of human-agent interaction needs special concerns that combine requirements from the perspectives of both the agents and the humans. The interactive behavior of such agents can be inspired by human-human interaction and can, additionally, be embodied by virtual agents or robots to scaffold intuitive interaction.
Robotics and Control (ROBOT)

Area Chairs: Nicola Basilico and Maria Gini

Topics:

    Coordination and collaboration in robotic systems
    Swarm and multi-robot collective behavior
    Robots in adversarial settings
    Perception and vision
    Networked systems and distributed robotics
    Foundation models for robotic agents
    Knowledge representation and reasoning in robotic systems
    Robot planning
    Mapping, localization and navigation
    Manipulation
    Robot control
    Robot learning
    Long-term (or lifelong) autonomy for robotic systems
    Execution monitoring and failure recovery for robots
    Robot modeling and simulation
    Explainability, trust and ethics for robots

Description:

Robots are embodied and situated agents with a deep connection to the physical world. Research on autonomous agents and multi-agent systems shares many challenges and synergies with robotics. We invite submissions that offer theoretical or applied contributions to agent-based approaches for single-robot and multi-robot systems, with particular emphasis on problems and solutions motivated by realistic scenarios typically encountered in robotic applications. All papers at the intersection of robotics and artificial intelligence, especially those grounded in agent research, fall within the scope of the Robotics and Control area of interest at AAMAS.
Innovative Applications (IA)

Area Chairs: Zehong (Jimmy) Cao and Gauthier Picard

Topics:

    Deployed or emerging applications of agent-based systems
    Realistic agent-based models of human organizations
    Evaluation of the cognitive capabilities of agent-based systems
    Integrated applications of agent-based and other technologies
    Challenges and best practices of real-world deployments of agent-based technologies

Description:

The innovative applications area aims to showcase successful applications and novel uses of agent-based technologies. We encourage research on emerging areas of agent-based applications with measurable benefits, on various topics such as (but not limited to) social good, sustainability, and ethical AI. The innovative applications area is keen to attract research that is not only triggered by real-world applications, but provides realistic beneficial solutions for these applications. Collaborations with relevant stakeholders is highly valued, as it helps demonstrate the feasibility and impact of the work.