Short-Term Gains and Emerging Long-Term Concerns: A Systematic Review and Synthesis of Meta-Analytic Evidence on AI's Impact on Student Creativity and Problem-Solving

Abstract

The integration of artificial intelligence (AI) in education presents a landscape of immediate opportunities and emerging long-term concerns. While many studies have documented the short-term effects of AI on student learning, a consolidated understanding of its sustained impact on core cognitive skills is lacking. This paper provides a systematic review of the current evidence, synthesizing findings from recent meta-analyses and longitudinal studies to evaluate the impact of AI on student creativity and problem-solving. Our review of the meta-analytic evidence, including data from over 50 experimental studies, confirms that AI-supported interventions can yield moderate short-term gains in creative thinking (SMD = 0.54). However, our synthesis of emerging longitudinal data and systematic reviews also surfaces significant concerns about the potential for cognitive over-reliance and the erosion of critical thinking skills over time. This paper maps the current evidence base, highlighting the robust quantitative findings on short-term impacts while calling attention to the critical, yet under-researched, long-term implications. We conclude that while the short-term benefits of AI are promising, they must be weighed against emerging long-term risks. We advocate for a balanced pedagogical approach that cultivates both AI literacy and enduring human skills, ensuring that technology serves as a scaffold for, rather than a substitute for, human ingenuity.