8th International Workshop on EXplainable, Trustworthy, and Responsible AI and Multi-Agent Systems
(EXTRAAMAS2026)

in conjunction with AAMAS 2026,
Cyprus, 25-26 May 2026

#Important Dates
Paper submission: 01/03/2026
Notification of acceptance: 25/03/2025
Early registration deadline: TBA
Workshop: 25-26/05/2026
Camera-ready (Springer post-proceedings): 10/06/2026
Submission link: https://easychair.org/conferences/?conf=extraamas2026



The International Workshop on EXplainable, Trustworthy, and Responsible AI and Multi-Agent Systems (EXTRAAMAS) has run since 2019 and has become a well-established forum at the intersection of Explainable AI (XAI), Agentic AI and Multi-Agent Systems (MAS). EXTRAAMAS focuses on explainability for agentic systems operating in dynamic, multi-stakeholder environments—where agents plan, negotiate, coordinate, and reason about norms. The workshop emphasizes the shift beyond static, post-hoc explanations toward interactive, context-aware, and evaluation-driven approaches, supporting systems that can explain-to-decide rather than merely explain-after. In its 8th edition, EXTRAAMAS 2026 continues to strengthen both foundational and applied research through four tracks spanning neuro-symbolic and hybrid approaches, explainable negotiation and conflict resolution, interactive explainability and LLM-based agentic systems, and legal/ethical perspectives.
The workshop is structured around four thematic tracks covering foundational, applied, interactive, and cross-disciplinary perspectives on explainable and trustworthy agentic AI, including symbolic and sub-symbolic approaches, negotiation and conflict resolution, interactive and LLM-based explainability, and legal and ethical dimensions.

The four tracks for this year are:

#Track 1: Foundations of Explainable and Agentic AI (Symbolic, Sub-symbolic, and Hybrid Approaches)
This track focuses on foundational approaches to explainability for agentic AI systems, spanning symbolic, sub-symbolic, and hybrid (neuro-symbolic) models. It addresses how explanations can be embedded within the reasoning cycle of autonomous agents, supporting planning, learning, coordination, and decision-making in complex environments.
Topics of interest include (but are not limited to):

- Explainable machine learning and neural networks
- Symbolic knowledge representation, injection, and extraction
- Neuro-symbolic and hybrid reasoning architectures
- Causal and counterfactual explanation models
- Surrogate models and abstraction techniques
- Explainable planning and decision-making
- Multi-agent architectures supporting explainability
- Evaluation and benchmarking of foundational XAI methods

#Track 2: Explainable Interaction, Negotiation, and Collective Decision-Making in Multi-Agent Systems
This track addresses explainability in interactive multi-agent settings, including negotiation, coordination, argumentation, and collective decision-making. As agents increasingly operate in open, human-facing and multi-stakeholder environments, transparent and trustworthy interaction mechanisms become essential for understanding, trust, and effective collaboration.
Topics of interest include (but are not limited to):
- Explainable negotiation protocols and strategies
- Explainable conflict resolution and coordination mechanisms
- Argumentation-based explanations of decisions and outcomes
- Explainable recommendation systems and preference learning
- Trustworthy voting and collective choice mechanisms
- User and agent profiling for transparency and accountability
- Human- and agent-centered evaluation studies
- Applications in robotics, IoT, virtual assistants, and socio-technical systems

#Track 3: Interactive, Conversational, and LLM-Based Explainable Agentic Systems
This track focuses on interactive and user-in-the-loop explainability, emphasizing dialogue, conversational interfaces, and adaptive interaction. It explores explainability challenges arising from LLM-based, tool-using, and hybrid agentic systems, including issues of reliability, faithfulness, and evaluation.
Topics of interest include (but are not limited to):
- Interactive and conversational explanation systems
- Explanatory dialogue and mixed-initiative interaction
- Context modeling and user modeling for explainability
- Prompt engineering and explanation-aware prompting
- Explainability challenges in LLM-based and tool-using agents
- Reliability, faithfulness, and hallucination mitigation
- Methodologies for evaluating interactive explanations
- Responsible and trustworthy deployment of LLM-driven agents

#Track 4: Explainable, Trustworthy, and Governed AI: Legal, Ethical, and Societal Perspectives
This track focuses on the legal, ethical, and societal dimensions of explainable and trustworthy AI,
addressing governance, accountability, and compliance in autonomous and agent-based systems deployed in sensitive and regulated domains.
Topics of interest include (but are not limited to):
- Explainability in AI & Law and legal reasoning systems
- Compliance-by-design and regulatory frameworks (e.g., EU AI Act)
- Fairness, bias mitigation, and transparency
- Accountability, liability, and auditability of AI systems
- Nudging, deception, and ethical design choices
- Normative reasoning and machine ethics
- Culture-aware and value-sensitive AI systems


#KEYNOTES
Keynote 1 Title: Evaluating Explanations in Multi-Agent and Human-AI Systems
Speaker: Prof. Sandip Sen, University of Tulsa
Abstract: This keynote discusses current challenges and methodologies for evaluating explanations in multi-agent systems and human–AI collaboration, highlighting limitations of existing metrics and future research directions.

Keynote 2 Title: Explainability Challenges for Agentic AI and User Rights
Speaker: Dr. Rachele Carli, Umeå University
Abstract: This keynote addresses the challenges posed by explainable AI models for agentic systems operating in complex socio-technical settings, with a focus on user rights, integrity, and accountability.


#Workshop Chairs
Prof. Dr. Davide Calvaresi, HES-SO, Switzerland
research areas: Real-Time Multi-Agent Systems, Explainable AI, BCT, eHealth,
mail: davide.calvaresi@hevs.ch

Dr. Amro Najjar, UNILU, Luxembourg
research areas: Multi-Agent Systems, Explainable AI, AI
mail: amro.najjar@uni.lu

Prof. Dr. Kary Främling, Umeå & Aalto University Sweden/Finland,
research areas: Explainable AI, Artificial Intelligence, Machine Learning, IoT
mail: Kary.Framling@cs.umu.se

Prof. Dr. Andrea Omicini
research areas: Artificial Intelligence, Multi-agent Systems, Soft. Engineering
mail: andrea.omicini@unibo.it, web page, Google Scholar


#Track Chairs
Dr. Giovanni Ciatto, University of Bologna, Italy – giovanni.ciatto@unibo.it
Prof. Rehyan Aydogan, Ozyegin University, Turkey – reyhan.aydogan@ozyegin.edu.tr
Rachele Carli, University of Bologna – rachele.carli2@unibo.it
Joris HULSTIJN: University of Luxembourg – joris.hulstijn@uni.lu

#Publicity Chair:
Elia Pacioni - HES-SO, Switzerland

#Advisory Board
Dr. Tim Miller, University of Melbourne
Prof. Dr. Leon van der Torre, UNILU
Prof. Dr. Virginia Dignum, Umea University
Prof. Dr. Michael Ignaz Schumacher

#Primary Contacts
Prof. Dr. Davide Calvaresi - davide.calvaresi@hevs.ch
Dr. Amro Najjar - amro.najjar@list.lu