"AI brain rot" and cognitive offloading Metacognition: knowing what you know

Heavier AI users scored lower on critical thinking in a UK survey, but this is a correlation

In an online survey of 666 people in the UK, those who said they used AI tools more also said they offloaded more thinking to devices, and they had lower critical-thinking scores. Younger participants used AI the most and scored lowest. The study measured everything at one point in time, so it cannot show that AI use caused the lower scores.

What they did

A 23-question questionnaire, built from items in the Halpern Critical Thinking Assessment and self-reported critical-thinking measures, asked about AI tool use, cognitive offloading (using devices for memory and problem-solving) and critical thinking, mostly on a 6-point scale. The author analysed it with ANOVA, correlations, multiple regression and random forest regression, and held semi-structured interviews with 50 participants.

Who took part

669 people in the UK recruited through social media posts (convenience and purposive sampling), of whom 666 were valid. Ages ran from 17 upwards; 110 were 17–25 and 86 were over 55.

What they found

• AI tool use and cognitive offloading went strongly together (r = +0.72).
• AI tool use and critical-thinking scores went strongly in opposite directions (r = −0.68).
• Cognitive offloading and critical thinking had the strongest negative link (r = −0.75), which the author reads as offloading being the route between AI use and lower critical thinking.
• Younger participants (17–25) reported the most AI use and offloading and had the lowest critical-thinking scores. More education went with higher critical-thinking scores.
• In interviews, many younger participants described relying on AI for tasks from looking things up to making decisions.

Effect size: Pearson correlations: +0.72 (AI use and offloading), −0.68 (AI use and critical thinking), −0.75 (offloading and critical thinking).

Limitations

• Cross-sectional and correlational: people who think less critically may simply use AI more, or a third factor (such as age or education) may drive both.
• Critical thinking was measured partly by self-report within the same survey, not by a full standalone test.
• A convenience sample recruited on social media in one country.
• Very strong correlations between survey scales can partly reflect how similar the questions are.
• The paper describes three age groups in its methods but reports five age bands in its tables.

What this means for learners and parents

• Treat this as a warning sign, not proof that AI harms thinking.
• What seems to matter is offloading: letting the tool do the remembering, judging and deciding. A family rule of "try first, then ask" targets exactly that.
• Older learners with more education scored higher regardless of AI use, which is a reminder that background knowledge still matters.

Put it into practice (free tools)

• Does AI make kids dumber? What the evidence says
• Who is doing the thinking? Check a prompt
• Family AI rules: try first, then ask

Citation

Michael Gerlich (2025). AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking. Societies 15(1), 6 (MDPI). https://doi.org/10.3390/soc15010006

Related summaries

• Unrestricted ChatGPT help raised practice scores but lowered exam scores · Hamsa Bastani et al., 2025 · Preregistered randomized controlled trial
• "Your Brain on ChatGPT": what the MIT essay-writing study did and didn't show · Nataliya Kosmyna et al., 2025 · Lab experiment with EEG
• ChatGPT made research feel easier but students' reasoning was weaker than with Google · Matthias Stadler et al., 2024 · Randomized experiment
• With AI, people scored better on reasoning puzzles but overestimated how well they did · Daniela Fernandes et al., 2026 · Two online experiments

Summary by the Ask Poppy team, AI-assisted and checked against the paper. Summarized 21 October 2026. Spotted an error? [email protected] · Editorial policy