Using Learning Theories to Evolve Human-Centered XAI: Future Perspectives and Challenges

2026-04-24 · Source: cs.AI updates on arXiv.org · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Human-Computer Interaction & Explainable AI · Depth: Expert, long

Summary

This paper proposes integrating learning theories into the Explainable AI (XAI) lifecycle to enhance human-centered XAI practices, particularly as AI models like GPT-3 grow exponentially in complexity and parameters. It argues that explanations primarily serve to foster learning, drawing parallels with human explanations that support understanding and adaptive behavior. The authors review various human learning theories—including behavioral, cognitivism, constructivism, experiential, humanistic, reflective, and social theories—and discuss how explanations influence learning at stages such as seeking, receiving, and producing them. The work identifies opportunities and challenges in using these theories to assess XAI needs, deliver learning experiences via explanations, and evaluate XAI methods, ultimately advocating for a shift from explanation-centric to learner-centric XAI to augment human skills and mitigate risks like over-reliance on AI.

Key takeaway

For AI scientists and research scientists developing XAI systems, you should consider adopting a learner-centered approach to explanation design. This means prioritizing how explanations facilitate human learning and skill augmentation over merely providing technically complete algorithmic details. By focusing on the "explainee" and integrating learning theories, you can mitigate risks like unwarranted trust or over-reliance on AI, ensuring explanations genuinely support human understanding and adaptive behavior rather than just satisfying curiosity.

Key insights

Infusing learning theories into XAI can shift focus from explanation complexity to human learning and agency.

Principles

• Explanations primarily foster learning.
• Learning is influenced by multiple cognitive processes.
• Learner-centered XAI enhances human agency.

Method

A learner-centered approach to XAI involves assessing needs using motivational and constructivist theories, designing explanations that trigger reflection, and evaluating methods based on educational and cognitive science principles.

In practice

• Use motivational theories to assess XAI needs.
• Design explanations to trigger reflection, not just inform.
• Evaluate XAI methods using learning objectives.

Topics

• Explainable AI
• Learning Theories
• Human-Centered AI
• AI Complexity
• Learner-Centric Explanations

Best for: AI Scientist, Research Scientist, AI Ethicist

Related on AIssential

• Explaining AI Without Code: A User Study on Explainable AI — Integrating XAI into no-code ML platforms improves usability and trust, but must balance novice accessibility with expert diagnostic depth.
• Where's the Structure? A Systematic Literature Review of Empirical Research on Human-AI Collaboration and Hybrid Intelligence for Learning — Effective human-AI collaboration for learning requires structured interaction, a gap identified in empirical research.
• A Definition of Good Explanations and the Challenges Explaining LLM Outputs — Good AI explanations combine counterfactuals with the interlocutor's prior beliefs to address LLM explainability challenges.
• We Need Explanation Cards to Connect Explanation Algorithms to the Real World — Explanation Cards bridge the gap between perceived and actual meaning of algorithmic explanations by providing crucial context and interpretation guidance.
• How to Leverage Explainable AI for Better Business Decisions — Explainable AI transforms business data into actionable insights by revealing the "why" behind predictions.
• Why Should I Trust You? (and Can I?) — AI models are black boxes because their complex internal reasoning is difficult to interpret, not because their underlying mechanics are unknown.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by cs.AI updates on arXiv.org.