📄 Abstract
Generative Artificial Intelligence (GenAI) is increasingly used to support learning and problem-solving, but its cognitive effects depend on how the support is designed. This integrative review examines how unrestricted, guided, scaffolded and collaborative forms of GenAI support influence cognitive engagement and independent problem-solving, with particular attention to managerial learning. A focused collection of 26 papers was screened, and 22 complete articles were retained for thematic synthesis. The evidence shows that unrestricted, complete-answer support can improve immediate task performance while increasing the risk of passive reliance, cognitive offloading and weaker retention. Guided, scaffolded and reflective forms of support are more likely to preserve active engagement because users must still interpret, question, verify and revise AI output. However, improved performance while AI is available does not necessarily show that the user can solve problems independently after the support is removed. Most available studies come from education and programming, while direct evidence from managerial and organizational settings remains limited. The review concludes that GenAI should support human reasoning rather than replace it, especially when the aim is to develop lasting judgement and independent problem-solving capability.
🏷️ Keywords
📚 How to Cite:
Rimsa Fathima Moulvi, A.P Muthulakshmi , GENERATIVE AI SUPPORT IN MANAGERIAL LEARNING: AN INTEGRATIVE REVIEW OF COGNITIVE ENGAGEMENT AND INDEPENDENT PROBLEM-SOLVING , Volume 12 , Issue 8, August 2026, EPRA International Journal of Multidisciplinary Research (IJMR) , Pages: 73 - 77 , DOI: https://doi.org/10.36713/epra31152