Exploring student and teacher perspectives on critical thinking with generative AI: early insights from the EduPartnerAI platform

Abstract The integration of artificial intelligence (AI) into education has raised important questions about its implications for students’ analytical reasoning and independent judgment. ChatGPT, a generative AI model capable of producing human-like responses, is widely used in academic contexts, yet its educational value requires careful evaluation. This paper presents and evaluates EduPartnerAI, a platform that integrates ChatGPT into student assignments and is designed to encourage critical engagement with AI-generated content. The platform records student-AI interactions, requires students to reflect on ChatGPT’s outputs, and provides teachers with tools to review observable indicators of student reasoning and AI response quality. An exploratory study involving 61 undergraduate students was conducted, relying on student reflection questionnaires, a usability survey, teacher evaluations, and an analysis of recorded interactions. Students generally perceived ChatGPT as a useful learning tool that provided clear explanations and practical examples, although some recognized that its responses could be overly general or incomplete. The teacher observed that students asked relevant questions and effectively incorporated AI-generated content into their assignments but experienced some difficulties with error detection, independent judgment, and critical evaluation. The teacher also judged ChatGPT’s outputs to be largely relevant, clear, and accurate, with few instances of bias or misleading information. Overall, the findings provide preliminary insights into student and teacher perspectives and observable patterns of engagement, highlighting both the educational potential of generative AI and the need for appropriate instructional scaffolding.