Artificial intelligence (AI)-assisted tools have become widely used to improve reading efficiency. This study aims to systematically investigate how AIGC-generated summaries modulate cognitive load and reading performance across varying levels of text difficulty, conceptualizing such summaries as digital nudges grounded in cognitive load theory and nudge theory.
This study used a 2 × 2 within-subjects design in which participants completed reading comprehension tasks under four conditions: easy versus difficult texts and the presence versus absence of AIGC-generated summaries. During the tasks, electroencephalography (EEG) signals, behavioral measures and questionnaire data were simultaneously collected.
Findings revealed that AIGC-generated summaries reduced theta/alpha ratio (TAR), shortened reading time, improved comprehension accuracy and alleviated subjective cognitive load across both levels of text difficulty. Interaction analyses further demonstrated that these load-reducing effects were more pronounced under high-difficulty conditions.
To the best of the authors’ knowledge, this study provides the first neurophysiological evidence for the cognitive benefits of AIGC-generated summaries in alleviating cognitive load during reading, thereby addressing limitations in prior behavior-based research. The findings not only offer practical guidance for the design of AI-assisted intelligent reading systems but also provide empirical evidence to support personalized summary delivery in libraries and digital reading platforms.
