Multi-stakeholder Perspective on Responsible Artificial Intelligence and Acceptability in Education
Project Overview
The document examines the role and impact of generative AI, particularly ChatGPT, in education from a multi-stakeholder perspective, underscoring its transformative potential across various educational contexts. It highlights key applications of AI, such as personalized learning, tutoring, and administrative support, while emphasizing critical factors influencing the perception of AI's utility, including agency, transparency, explainability, and privacy. The findings reveal significant variations in acceptance and trust in AI among stakeholders—students, teachers, and parents—indicating the necessity for a nuanced approach to understanding AI's role and ethical implications in educational environments. Ultimately, the document calls for a comprehensive framework that addresses these factors to enhance the integration of generative AI in education and foster greater confidence and acceptance among all involved parties.
Key Applications
Feedback-AI
Context: High school settings for exam grading and personalized tutoring
Implementation: AI technology that marks students' exams, provides feedback, and adjusts content difficulty to engage students effectively.
Outcomes: ['Reported issues of trust related to agency and fairness in grading', 'Potential for enhanced personalized learning experiences']
Challenges: ['Concerns about AI agency leading to lower confidence and perceived justice', 'Need for balance between AI supervision and independence']
Support-AI
Context: Chatbot answering students' questions and providing emotional support outside class hours
Implementation: Conversational agent based on AI algorithms capable of understanding and responding to both informational and emotional needs.
Outcomes: ['Enhanced accessibility to information for students', "Support for students' emotional well-being"]
Challenges: ['Privacy concerns regarding data sharing with teachers', 'Concerns about privacy in handling sensitive emotional data']
Implementation Barriers
Ethical
Concerns about data privacy, autonomy of AI systems, and ethical use in education.
Proposed Solutions: Implement robust data privacy measures and clear ethical guidelines for AI use.
Trust
Lack of trust in AI systems due to perceived agency, transparency issues, and the need for enhanced explainability.
Proposed Solutions: Enhance explainability and transparency of AI processes to build stakeholder trust.
Implementation
Challenges in integrating AI into existing pedagogical frameworks while maintaining educational integrity and agency.
Proposed Solutions: Carefully calibrate AI's role to ensure it aligns with educational goals.
Project Team
A. J. Karran
Researcher
P. Charland
Researcher
J-T. Martineau
Researcher
A. Ortiz de Guinea Lopez de Arana
Researcher
AM. Lesage
Researcher
S. Senecal
Researcher
P-M. Leger
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: A. J. Karran, P. Charland, J-T. Martineau, A. Ortiz de Guinea Lopez de Arana, AM. Lesage, S. Senecal, P-M. Leger
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai