Evidence Library · AI & Cognitive Demand

LLM-assisted essay writing shows weaker EEG connectivity and lower reported ownership in a small preprint study

Preprint or emerging researchPreprint

Emerging, unreviewed finding · Watchlist · 2025

LLM-assisted essay writing shows weaker EEG connectivity and lower reported ownership in a small preprint study

In a preprint, 54 adults complete three essay-writing sessions using an LLM, web search, or no external tool. The LLM group shows the weakest EEG connectivity and reports the lowest ownership of its essays. Only 18 participants complete the fourth crossover session.

Research basis

Small repeated-session EEG experiment with a crossover session. 54 adults completed the first three sessions; 18 completed the fourth crossover session.

What it supports

A preliminary task-specific signal that different writing tools were associated with different neural-connectivity, recall, and ownership patterns.

Boundary

This is a small task-specific study. It does not demonstrate brain damage, generalized cognitive decline, long-term impairment, or effects across workplace tasks. Caution: The study is a preprint and not yet peer-reviewed or published.

BOL relevance

A hypothesis-generating AI and cognition signal that should remain separate from the foundational BOL evidence spine.

Source class

Preprint or emerging research

Role in BOL

Watchlist

Population context

General adult population

Evidence basis

EEG, task performance, recall, and self-reported ownership

Integrity notes

A separate methodological critique is also a preprint and has not been peer reviewed.

Publication version: Version 2.

Publication status last checked 26 July 2026.

No peer-reviewed journal article identified.

LLMEEGauthorshippreprint

Source record checked 26 July 2026.

This entry is part of the Bio-Optimized Leadership™ Evidence Library. Each claim states its research basis, what the source supports, what it does not support, and how it relates to the BOL model.