Learning Economics In The Era Of Chatgpt: From Deep Learning To Critical Thinker

Abstract

Purpose: This article aims to examine systematic literature review (SLR) examines how a deep learning approach can reposition students from passive AI users toward critical thinkers in economics learning.

Design/Methodology/Approach: The review synthesizes peer-reviewed literature published during 2022–2026, emphasizing studies indexed or published in journals covered by Scopus and/or Web of Science and directly addressing GenAI, ChatGPT, higher education, critical thinking, self-regulated learning, assessment, and AI literacy. A PRISMA-oriented screening and thematic synthesis framework was used to organize the evidence.

Findings: The synthesis indicates that ChatGPT produces the greatest educational value when embedded in inquiry-based, problem-based, reflective, dialogical, and evidence-verification activities. Deep learning functions as a pedagogical filter: students are required to question AI outputs, verify claims against authoritative data and scholarly sources, compare alternative economic explanations, articulate assumptions, and formulate independent conclusions. The review further identifies AI literacy, metacognitive regulation, assessment redesign, transparency, and lecturer scaffolding as key conditions for successful integration. The proposed transformation model moves students through four stages: AI User, AI-Assisted Learner, Critical AI User, and Critical Thinker.

Research Limitations/Implications: This article is based on secondary literature and does not present primary empirical evidence from specific students. Future research should employ quantitative, qualitative, or mixed-methods approaches that involve students.

Originality/Value: This study offers an original perspective on economics education in the era of ChatGPT by examining how deep learning approaches can be integrated with generative artificial intelligence to foster students’ critical thinking skills. Unlike studies that primarily focus on the effectiveness or efficiency of AI-assisted learning, this article emphasizes the importance of moving beyond technology-assisted knowledge acquisition toward meaningful learning, critical inquiry, analytical reasoning, and reflective judgment. Its value lies in proposing a conceptual framework that positions ChatGPT as a learning facilitator rather than a substitute for students’ independent thinking, particularly in analyzing economic issues, evaluating information, and making evidence-based decisions. This perspective contributes to the development of more critical, reflective, and responsible economics learning practices in the age of generative AI.

Paper Type: Literature Review