Abstract:

The rapid integration of Artificial Intelligence (AI) into educational contexts is disrupting long standing assumptions about teaching, learning, cognition, authorship, and assessment. What was once a relatively stable pedagogical relationship—teacher as transmitter of knowledge, student as recipient, and assessment as evaluation of individual mastery—has been fundamentally unsettled by systems capable of generating fluent text, solving complex problems, composing code, and simulating human like reasoning. This transformation alters not merely classroom practice but the conditions under which knowledge is produced and evaluated.

Educational assessment has historically relied on the assumption that student outputs reliably reflect internal understanding. In AI mediated environments, this link is increasingly fragile: sophisticated artefacts may be produced with minimal conceptual engagement, while deep intellectual work may occur through dialogue with machines yet remain invisible in the final product. This paper argues that education has therefore entered a new epistemic phase in which cognition is partially distributed across human and machine systems, requiring educators to move beyond evaluating products toward diagnosing thinking processes—how learners frame questions, test assumptions, select information, integrate perspectives, and exercise judgement.

The paper proposes a redefinition of academic intelligence suited to this context, highlighting emerging competencies such as prompting competence, critical selection among machine generated outputs, synthesis across sources and disciplines, metacognitive regulation of AI reliance, epistemic humility, and ethical judgement. These capacities, it argues, constitute a new cognitive literacy essential for academic integrity, professional competence, and democratic participation.

Drawing on the historical caution embedded in Lamarck’s theory of use and disuse, the paper further warns that sustained over reliance on AI may weaken independent reasoning, attentional endurance, and tolerance for intellectual struggle. While AI can function as a powerful educational resource, it may also become a cognitive surrogate that reshapes thinking toward speed and fluency at the expense of depth and autonomy.

The paper advances the normative claim that AI must be positioned as a scaffold for expanded human reasoning rather than a substitute for intellectual effort. It concludes that teacher education and assessment systems must be fundamentally reoriented to evaluate cognitive processes, cultivate metacognition, and preserve human judgement in an age of augmented minds.

Keywords:

Artificial Intelligence in Education, Assessment Reform, Cognitive Processes, Distributed Cognition, Metacognition, Teacher Education, Ethics of AI, Human–Machine Collaboration, Learning Theory.