Institutions of higher education are rapidly incorporating artificial intelligence into their programs. A study of 65 R1 universities revealed that 63% are actively promoting the use of generative AI, with many offering detailed guidance for its integration into teaching (McDonald et al., 2025). The expectation is that AI will enhance student thinking, customize learning, and better equip graduates for a tech-driven job market. However, emerging research presents a more nuanced picture, which educators and academic leaders must consider.
The assumption that AI enhances learning is driving its adoption, but the evidence is mixed and often contradicts this belief. A Swiss study showed that frequent use of AI tools negatively impacts critical thinking, as students relying on AI for cognitive tasks showed diminished critical thinking skills (Gerlich, 2025). Researchers at Wharton identified a phenomenon called “cognitive surrender,” where participants accepted AI-generated outputs without scrutiny, transferring decision-making to machines (Shaw & Nave, 2026). A 2025 meta-analysis of 18 studies on generative AI found that over-reliance on these tools erodes higher-order thinking skills like critical analysis and problem-solving (Qu et al., 2025).
The issue is not confined to AI alone. Since the widespread adoption of laptops and tablets in K–12 education in the early 2000s, there has been an unprecedented decline in IQ scores. International assessments such as PISA, TIMSS, and PIRLS have linked this decline to increased technology use (Horvath, 2026; Rogelberg, 2026). A 2025 randomized controlled trial highlighted that students using ChatGPT as a study aid retained significantly less knowledge 45 days after instruction compared to those who did not use it (Barcaui, 2025).
Effective learning requires deep engagement, productive struggle, and repetition. When AI is used to summarize readings, draft essays, or guide problem-solving, students may produce polished work but fail to develop lasting knowledge or skills. The shortcut bypasses the learning process. In teacher education courses, having students create lesson plans themselves fosters engagement with real questions and learning through effort, unlike analyzing AI-generated plans.
AI’s role in higher education should be carefully defined, prioritizing student learning over engagement, efficiency, or career readiness. Faculty should default to “offline pedagogy,” integrating AI only when it genuinely supports learning conditions. These include reinforcing content knowledge, fostering deep processing, encouraging independent thinking, and promoting meaningful human interaction.
Four critical questions can guide decisions on AI integration: Does the AI tool help students engage with core disciplinary content? Does it facilitate the application of learning to new contexts? Does it support independent, evidence-based reasoning? Does it preserve meaningful human interaction? These questions form a framework for evaluating AI’s role in education.
Resistance to AI is impractical, yet thoughtful integration is necessary. While AI can enhance learning in certain contexts, its rapid adoption often outpaces evidence of its effectiveness. Educators face pressure to use tools endorsed by their institutions and popular with students, despite unclear benefits to learning. It’s crucial to remain cautious, recognizing that AI developers may prioritize profit over educational outcomes. The outlined questions offer a foundational approach for assessing AI’s impact on learning.
Original Source: facultyfocus.com
