Abstract:

The rapid integration of generative artificial intelligence in higher education has largely been framed as a challenge to academic integrity or a shortcut to cognitive effort. This paper proposes an alternative framing: AI (Artificial Intelligence) as a cognitive prosthetic. Rather than positioning AI tools as substitutes for thinking, this study investigates their potential role as executive-function scaffolds for students experiencing attention regulation difficulties, task initiation barriers, and cognitive overload.

Grounded in cognitive load theory and inclusive education design, the research examines how AI can function as structured planning support, externalized working memory, and metacognitive prompting. Through guided implementation in undergraduate coursework, students were encouraged to use AI primarily for idea organization, task breakdown, clarification of complex instructions, and reflective questioning rather than content generation.

Preliminary classroom findings indicate that when pedagogically framed and transparently integrated, AI may reduce cognitive friction without diminishing analytical depth. Students reported improved task persistence and clarity of thought, while assessment outcomes showed no reduction in critical reasoning quality. The paper argues that intentional AI integration can expand accessibility and support neurodivergent learners, provided its role remains scaffolded rather than substitutive. This reframing positions AI not as a threat to intellectual rigor or academic integrity, but as a potential assistive technology within inclusive higher education ecosystems.

Keywords:

Artificial Intelligence, Neurodivergent, Cognitive load theory, Pedagogy, academic integrity.