Generative Artificial Intelligence as a Metacognitive Partner: a Theoretical Model of Shared Regulation of Learning in Higher Education
TL;DR
The article proposes the Model of Metacognitive Coregulation with Generative Artificial Intelligence, which explains under what conditions this technology can operate as an external agent of shared regulation that scaffolds student planning, monitoring and evaluation instead of replacing them.
Abstract
The rapid adoption of generative artificial intelligence in higher education has shifted the debate from academic integrity toward a deeper question: how this technology transforms the processes through which students regulate their own learning. Recent evidence shows ambivalent results, since artificial intelligence assistance improves immediate performance but can induce metacognitive laziness, cognitive offloading and dependence. This conceptual article, designed as a theory synthesis, integrates three research streams: regulation of learning and metacognition, social modes of regulation, and human interaction with generative artificial intelligence. As a result, the article proposes the Model of Metacognitive Coregulation with Generative Artificial Intelligence, which explains under what conditions this technology can operate as an external agent of shared regulation that scaffolds student planning, monitoring and evaluation instead of replacing them. The model is formalized in eight testable propositions and a concept map, and concludes with an empirical research agenda for higher education.