AI Makes You Smarter but None the Wiser

§ 01Overview

BackTier Analysis — 2026

Story at a Glance

• A peer-reviewed study in *Computers in Human Behavior* (vol. 175, 2026) from Aalto University, LMU Munich and partners found that people using AI score higher on logical-reasoning tasks — and overestimate how well they did.
• In Study 1 (N = 246), AI users gained about three points on LSAT reasoning problems but overestimated their score by about four.
• Higher AI literacy predicted *worse* self-assessment. Study 2 (N = 452) replicated the result.

§ 02What the Study Found

The researchers asked participants to solve 20 Law School Admission Test logical-reasoning problems with AI help, then estimate their own score. AI lifted performance. It lifted confidence further.

The usual Dunning–Kruger pattern — weak performers overestimating, strong performers underestimating — disappeared. With AI in the loop, almost everyone overestimated. AI levelled performance and self-assessment at the same time.

§ 03Why It Matters for AI Visibility

This is The Lisa Latte Effect measured in a lab: AI output feels like your own competence.

For businesses, the same dynamic shows up in AI visibility work. A team asks ChatGPT about itself, gets a fluent answer, and assumes the machine understands the entity. Fluency is not accuracy. The only reliable check is external measurement: what models actually cite, include and select across repeated, dated queries.

Practical Rules

• Measure, don't self-assess. Track citations and selections over time instead of trusting one good answer.
• Treat AI literacy as a risk factor for overconfidence, not a guarantee of judgement.
• Build verification into the workflow — sources, dates and a second check before anything is published.

§ 04Sources

• Fernandes, D., Villa, S., Nicholls, S., Haavisto, O., Buschek, D., Schmidt, A., Kosch, T., Shen, C., & Welsch, R. (2026). AI makes you smarter but none the wiser: The disconnect between performance and metacognition. *Computers in Human Behavior*, 175, 108779. https://doi.org/10.1016/j.chb.2025.108779
• Aalto University news: AI use makes us overestimate our cognitive performance. https://www.aalto.fi/en/news/ai-use-makes-us-overestimate-our-cognitive-performance