Multi-Agent Systems in Education: A Survey from the Trustworthiness Perspective

Chahana Dahal; Jinming Chen; Muchao Ye; Zuobin Xiong

doi:10.1109/TAI.2026.3695502

Back

Multi-Agent Systems in Education: A Survey from the Trustworthiness Perspective

Journal article

Peer reviewed

Multi-Agent Systems in Education: A Survey from the Trustworthiness Perspective

Chahana Dahal, Jinming Chen, Muchao Ye and Zuobin Xiong

IEEE transactions on artificial intelligence

05/21/2026

DOI: 10.1109/TAI.2026.3695502

View Online

Abstract

Multi-agent systems are becoming an important part of educational technology by helping students learn through collaboration, feedback, and adaptive support. This paper reviews previous and recent research on multi-agent learning systems and explores the roles agents can take, such as tutors, peers, or facilitators. To the best of our knowledge, this is the first comprehensive survey of multi-agent learning systems that introduces a pedagogy-based role and interaction taxonomy, compares system architectures beyond LLMs, and links design to learning theory. The fields are grouped on the basis of system architecture, learning scenarios, and connections to learning theories such as constructivism and the Zone of Proximal Development. The survey examines how these systems are evaluated and the challenges of measuring learning, engagement, and collaboration. In addition, this work provides a comprehensive study of emerging trustworthiness frameworks that address safety, privacy, fairness, and transparency in multi-agent learning systems. Finally, the paper concludes with open challenges and future directions in using the multi-agent learning system in education scenarios, including the need for better simulation environments, stronger coordination between humans and AI agents, etc.

Safety

AI Agents

AI in Education

Artificial intelligence

Cognition

Cognitive systems

Design methodology

Feedback

Large language models

Learning (artificial intelligence)

Modeling

Multi-agent systems

Trustworthy AI

Details

Title: Subtitle: Multi-Agent Systems in Education: A Survey from the Trustworthiness Perspective
Creators: Chahana Dahal - University of Nevada, Las Vegas
Jinming Chen - independent researcher
Muchao Ye - University of Iowa
Zuobin Xiong - University of Nevada, Las Vegas
Resource Type: Journal article
Publication Details: IEEE transactions on artificial intelligence
DOI: 10.1109/TAI.2026.3695502
ISSN: 2691-4581
eISSN: 2691-4581
Publisher: IEEE
Number of pages: 16
Language: English
Electronic publication date: 05/21/2026
Academic Unit: Computer Science
Record Identifier: 9985166827702771

Metrics

1 Record Views