The Neurocognitive Implications of Using ChatGPT in the Writing Process by Young Children

The rapid adoption of ChatGPT in education has sparked both enthusiasm and concern, particularly regarding its impact on the cognitive aspects of writing. Despite its potential to personalize learning and provide scalable writing support, little is known about ChatGPT’s effects on thinking skills that depend on writing and writing skills that depend on thinking. This literature review examines the neurocognitive implications of integrating ChatGPT into the writing processes of children aged 5–11, synthesizing evidence from 216 sources through a transdisciplinary lens informed by fields including language, technology, education, cognitive neuroscience, and developmental psychology. Writing is presented as a meta-skill essential for cognitive development across subjects. Evidence shows that AI writing assistants can provide personalized scaffolding, expand feedback, and strengthen metacognitive engagement. However, risks include cognitive offloading, diminished student voice, and bypassing critical stages in cognitive development, such as planning and revision. The review reveals that ChatGPT's impact appears to be contextually dependent rather than inherently beneficial or harmful, with effects varying across different stages of the writing process. A stage-by-stage analysis demonstrates that the same AI features can either support or hinder cognitive development depending on timing, instructional context, and developmental readiness. Findings suggest balancing benefits and risks depends on three elements—instructional design, adult guidance, and AI literacy—working together to preserve the developmental cognitive goals of writing education. This paper proposes preliminary, evidence-informed criteria for developmentally appropriate integration and outlines stakeholder roles for educators, parents, policymakers, researchers, students, and technology developers. As the long-term cognitive effects of sustained AI-supported writing cannot yet be measured due to its newness, research needs include readiness for prompting, children’s ability to detect AI bias, and diminished rehearsal of important cognitive functions upon which higher order cognition is developed. The review calls for skeptical optimism to leverage ChatGPT’s potential while applying safeguards grounded in writing pedagogy and child development. Immediate, coordinated action paired with ongoing research is essential to ensure AI integration strengthens learning, safeguards cognitive growth, and prepares children for a future where AI is integral to daily life.