GenAI's effects on higher-order thinking are mixed: support for reflection and idea expansion alongside cognitive offloading and no overall innovativeness gain¶

Claim · All claims
Evidence · 1 study (2 entries) · review r2 · q2–q3
Currency · newest study 2026 · checked against its sources 2026-10-20 (automated, 2 of 2 entries) · not reviewed by a person

Subclaims¶

q2 i? Reviewed studies describe GenAI supporting reflective thinking, pattern-finding and devil's-advocate challenge, but also report reduced critical engagement and homogenised ideas when over-relied on. → Wei 2026 q3 i? In a randomised ideation experiment, the AI collaborator produced more innovative suggestions but did not significantly enhance participants' overall innovativeness. → Wei 2026 (2)

Evidence¶

Wei 2026¶

Wei, Y., and Perkins, M. (2026). Generative AI and student collaboration: a scoping review of group work processes, outcomes, and risks. International Journal for Educational Integrity. https://doi.org/10.1007/s40979-026-00217-x

q2 · i? · review · r2

The review reports, citing Baskara, Patole and He et al., perceived cognitive offloading risks; the review notes only two empirical investigations (He et al. 2024; Scala et al. 2025) support critical-thinking concerns.

"Overreliance on GenAI might demotivate learners from thinking critically (Baskara 2024; Patole 2024), and participants expressed concern that "I'm not critically thinking" (He et al. 2024) when delegating problem-solving tasks to AI, highlighting perceived risks of cognitive offloading."

Wei 2026 (2)¶

Wei, Y., and Perkins, M. (2026). Generative AI and student collaboration: a scoping review of group work processes, outcomes, and risks. International Journal for Educational Integrity. https://doi.org/10.1007/s40979-026-00217-x

q3 · i? · review · r2

The review reports Scala et al.'s (2025) randomised experiment comparing human-only ideation with AI-supported teams, in which the AI collaborator "produced more innovative suggestions" without significantly enhancing participants' overall innovativeness; equivalence was not tested.

"a randomised experiment comparing human-only ideation with AI-supported teams revealed that while the AI collaborator produced more innovative suggestions, it did not significantly enhance participants' overall innovativeness."

Discussion¶

Learner Variables¶

• Critical Thinking — outcome: instruction changes it

Related Claims¶

• Dependent versus autonomous cognitive offloading to GenAI shows opposite motivational and cognitive outcome patterns — a narrower finding that bears on this claim
• AI tool use is negatively correlated with critical thinking, mediated by cognitive offloading and strongest in younger users (review attribution to Gerlich) — a narrower finding that bears on this claim
• GenAI learning outcomes depend more on use quality (capability, reflective engagement, retained agency) than on frequency of use — a broader claim this one bears on
• Metacognitive self-regulation mediates the relationships between all three AI interaction mechanisms and both higher-order outcomes — related
• Need for cognition correlates positively with critical genAI use and negatively with perceived usefulness of genAI — related
• The Struggling profile exhibited a fragmented AI-support pattern, with strong ENA connections between foundational scaffolding and irrelevant/off-task responses alongside cognitive offloading and collaborative impasse — related
• Systematic review of 43 investigations finds no consistent evidence that higher education produces more critical analytic ways of thinking — related
• Students’ actual genAI use correlates with interest and perceived usefulness of genAI but not with knowledge or critical-thinking dispositions — related
• Perceived learning benefits are higher when GenAI augments students' capabilities and lower when it hurts critical thinking or causes stress — related
• Cognitive offloading to generative AI is associated with reduced learner engagement, lower critical thinking, and lower-quality reasoning — related
• GenAI shows a dual-mechanism profile: meta-analytic learning benefits coexist with dependence-related costs to critical thinking — related
• Aggregate interaction volume did not demonstrate reduced human input; the strongest observed pattern was capacity expansion rather than proven labor substitution — related