Cognitive Offloading

A lens, not a label.

AI can help us reach further. But when it becomes our first response to every question, we may miss opportunities to practise the thinking that makes us human.

Keep the evidence in view.

Microsoft Research and Carnegie Mellon (CHI 2025) reported an association between higher confidence in AI and less critical-thinking activity. A 2025 MIT Media Lab EEG preprint reported weaker neural connectivity, recall and ownership among LLM-assisted writers. Cognitive offloading is an established concept; these AI-specific findings remain emerging, not a conclusion that every use of AI weakens thinking. [3,4]

Observation, without a label.

These patterns are described in the executive report. They do not establish a diagnosis or prove that AI is the cause.

• 01Starting with AI rather than an unaided attempt.
• 02Poor recall of AI-assisted work or difficulty explaining the reasoning.
• 03Accepting fluent output without verification.
• 04Falling confidence when working without AI.

Trace the source.

[1] Julia Worrall, RN. “AI Hyperarousal Syndrome (AIHS) — Clinical Definition & Priority Record.” Public clinical page, accessed 19 August 2026.

[3] Lee HP, Sarkar A, Tankelevitch L, et al. “The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers.” CHI 2025.

[4] Kosmyna N, Hauptmann E, Yuan YT, et al. “Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task.” MIT Media Lab / arXiv preprint, 2025.

[12] Buçinca Z, Malaya MB, Gajos KZ. “To Trust or to Think: Cognitive Forcing Functions Can Reduce Overreliance on AI in AI-assisted Decision-making.” 2021.

Make room for the human.

• 01Form your own answer before asking AI.
• 02Recall what you know before looking it up.
• 03Verify outputs and integrate them with your own judgment.
• 04Keep time for writing, learning and problem-solving without AI.