Detecting LLM-Assisted Academic Dishonesty using Keystroke Dynamics

TL;DR

This paper advances keystroke dynamics as a behavioral method to detect AI-assisted academic dishonesty, outperforming text-based detectors especially in realistic and adversarial scenarios.

Contribution

It extends prior work by expanding data, formalizing threat models, and empirically comparing keystroke-based detection with text-only methods and humans.

Findings

Keystroke-based models outperform text-only detectors in practical scenarios.

Detection performance decreases under adversarial conditions.

Expanded dataset with 130 participants and explicit paraphrasing condition.

Abstract

The rapid adoption of generative AI tools has heightened concerns regarding academic integrity, as students increasingly engage in dishonest practices by copying or paraphrasing AI-generated content. Existing plagiarism detection systems, which rely primarily on text-intrinsic features, are ineffective at identifying AI-assisted or paraphrased submissions. Our prior conference work introduced a behavioral detection approach that leverages how text is produced, captured through keystroke dynamics, in addition to what is written, enabling discrimination between genuine and assisted writing. That study, conducted on keystroke data from 40 participants, demonstrated promising performance. This paper substantially extends and systemizes the prior work by: (1) expanding the dataset with 90 additional participants and introducing an explicit paraphrasing condition to model realistic plagiarism strategies; (2) formalizing a threat model and evaluating detection under adversarial and deception-oriented scenarios; and (3) performing a comprehensive empirical comparison against state-of-the-art text-only detectors and human evaluators. Experimental results demonstrate that keystroke-based models significantly outperform text-based approaches in practical deployment settings, while revealing limitations under more challenging adversarial conditions.