Work

Cognitive Augmentation or Cognitive Offloading? Artificial Intelligence in Education and Lifelong Learning

Year: 2026
Type: article
Abstract: Research on artificial intelligence (AI) and education has developed across largely separate domains, including learning outcomes, academic integrity, institutional governance, and lifelong-learning demand. This study integrates these domains through a common framework distinguishing cognitive augmentation, where AI supports human thinking, from cognitive offloading, where it substitutes for it. It synthesizes 2025–2026 evidence from the OECD, UNESCO, MIT, RAND Corporation, UC Berkeley, HEPI, Gallup, the Walton Family Foundation, official Türkiye sources, and triangulated market data. PISA 2025 findings covering approximately 760,000 students across 91 countries and economies provide cross-national evidence: students who did not use AI to draft assignments outperformed those who did, while frequent users trained to critically evaluate AI-generated information performed better than those without such training. Additional evidence shows short-term AI-supported performance gains without equivalent retained learning, discipline-specific patterns of AI-assisted academic misconduct, rapidly increasing AI-generated coursework, and widespread teacher AI use despite limited formal governance. Türkiye is examined as a policy case linking economy-wide and education-specific AI strategies. Overall, the findings indicate that AI’s educational impact depends less on the technology itself than on pedagogical design, assessment redesign, and institutional governance capacity.
Institution: Marmara University
Language: English
Subfield: Education
Open Access status: hybrid
OpenAlex ID: W7221539505
FWCI:
Cites: 8
Cited by:
Related to: 10