The Path to Conversational AI Tutors: Integrating Tutoring Best Practices and Targeted Technologies to Produce Scalable AI Agents

2026-02-22

AI Summary

This work addresses the challenge of designing conversational AI tutors that are both educationally effective and scalable, capable of emulating high-quality human tutoring while supporting real-time student interaction. Integrating insights from human tutoring research, intelligent tutoring systems, and generative AI, the study proposes a four-dimensional design framework—Retain, Adapt, Focus, and Investigate. The framework retains proven mechanisms such as knowledge tracing, leverages generative AI for dynamic content generation and dialogic scaffolding, emphasizes student agency and deep cognitive diagnosis, and incorporates socio-emotional dimensions. This approach establishes a viable pathway for developing AI tutors that balance instructional efficacy, scalability, and learner motivation, laying the groundwork for future research on effectiveness validation, user experience, and human-AI collaborative teaching.

Technology Category

• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Intelligent Robots: Embodied AI

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: Uses of LLMs and GenAI for marketplace design, bidding, and strategic interactions

Abstract

The emergence of generative AI has accelerated the development of conversational tutoring systems that interact with students through natural language dialogue. Unlike prior intelligent tutoring systems (ITS), which largely function as adaptive and interactive problem sets with feedback and hints, conversational tutors hold the potential to simulate high-quality human tutoring by engaging with students' thoughts, questions, and misconceptions in real time. While some previous ITS, such as AutoTutor, could respond conversationally, they were expensive to author and lacked a full range of conversational ability. Generative AI has changed the capacity of ITS to engage conversationally. However, realizing the full potential of conversational tutors requires careful consideration of what research on human tutoring and ITS has already established, while also unpacking what new research will be needed. This paper synthesizes tenets of successful human tutoring, lessons learned from legacy ITS, and emerging work on conversational AI tutors. We use a keep, change, center, study framework for guiding the design of conversational tutoring. We argue that systems should keep proven methods from prior ITS, such as knowledge tracing and affect detection; change how tutoring is delivered by leveraging generative AI for dynamic content generation and dialogic scaffolding; and center opportunities for meaning-making, student agency, and granular diagnosis of reasoning. Finally, we identify areas requiring further study, including efficacy testing, student experience, and integration with human instruction. By synthesizing insights from human tutoring, legacy ITS, and emerging generative AI technologies, this paper outlines a research agenda for developing conversational tutors that are scalable, pedagogically effective, and responsive to the social and motivational dimensions of learning.

Problem

• Conversational AI Tutors
• Intelligent Tutoring Systems
• Generative AI
• Human Tutoring
• Scalable AI Agents

Innovation

• Conversational AI Tutors
• Generative AI
• Dialogic Scaffolding
• Knowledge Tracing
• Student Agency