aiqbench analysis
This research demonstrates that generative AI leaves distinct temporal signatures in keystroke dynamics, allowing classifiers to nearly perfectly distinguish between authentic collaboration and wholesale delegation across both writing and coding tasks. The study reveals a critical divergence in how humans interact with text versus code generators, where code suggestions are accepted far less frequently than text insertions, suggesting different modes of overreliance. These findings provide a robust evidentiary basis for detecting academic dishonesty without relying on ethically contentious post-hoc detectors that analyze only final artifacts.
Does AI Assistance Leave a Temporal Fingerprint? Detecting Overreliance in AI-Assisted Writing and Programming
arXiv:2609.17883v1 Announce Type: cross Abstract: The rapid adoption of generative AI has made final artifacts unreliable evidence of student learning, and AI detectors that examine only the finished product are inaccurate and ethically contentious. Process data offers an alternative, but prior work covers only English essay writing. We ask whether AI assistance carries a temporal signature, whether it generalizes from writing to programming, and whether it distinguishes ordinary collaboration from wholesale delegation. We analyze three public corpora: CoAuthor (1,447 keystroke-level co-writing sessions), RealHumanEval (editor telemetry from 243 programmer records), and a pre-LLM CS1 corpus (5.1 million keystrokes) as a human-only baseline, comparing minimal-AI work, collaborative AI use, and simulated wholesale delegation. Three findings emerge. First, the signature generalizes: AI contributions arrive in bursts far outside the author's own baseline in both mediums (paired d_z = 1.13 and 3.54). Second, engagement diverges by medium: 93% of AI-inserted characters survived to writers' final documents, while only 14% of accepted code suggestions survived intact. Third, classifiers using only observable temporal features separate simulated delegation from authentic work nearly perfectly (F1 $\geq$ 0.997; at most 0.5% of real work misclassified), while ordinary collaboration remains hard to distinguish from unassisted work. Temporal evidence flags wholesale delegation rather than assistance, positioning process visibility as a candidate evidentiary basis for academic integrity, pending validation in authentic coursework.
Title:Does AI Assistance Leave a Temporal Fingerprint? Detecting Overreliance in AI-Assisted Writing and Programming
Abstract:The rapid adoption of generative AI has made final artifacts unreliable evidence of student learning, and AI detectors that examine only the finished product are inaccurate and ethically contentious. Process data offers an alternative, but prior work covers only English essay writing. We ask whether AI assistance carries a temporal signature, whether it generalizes from writing to programming, and whether it distinguishes ordinary collaboration from wholesale delegation. We analyze three public corpora: CoAuthor (1,447 keystroke-level co-writing sessions), RealHumanEval (editor telemetry from 243 programmer records), and a pre-LLM CS1 corpus (5.1 million keystrokes) as a human-only baseline, comparing minimal-AI work, collaborative AI use, and simulated wholesale delegation. Three findings emerge. First, the signature generalizes: AI contributions arrive in bursts far outside the author's own baseline in both mediums (paired d_z = 1.13 and 3.54). Second, engagement diverges by medium: 93% of AI-inserted characters survived to writers' final documents, while only 14% of accepted code suggestions survived intact. Third, classifiers using only observable temporal features separate simulated delegation from authentic work nearly perfectly (F1 $\geq$ 0.997; at most 0.5% of real work misclassified), while ordinary collaboration remains hard to distinguish from unassisted work. Temporal evidence flags wholesale delegation rather than assistance, positioning process visibility as a candidate evidentiary basis for academic integrity, pending validation in authentic coursework.
Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
arXiv:2609.17883 [cs.HC]
(or arXiv:2609.17883v1 [cs.HC] for this version)
https://doi.org/10.48550/arXiv.2609.17883 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 326 words
Related coverage
2026.8: Approachable by design
Home Assistant · Aug 5, 2026
Links #3: 2026/06 Part 1
LessWrong · Jun 15, 2026
Build a Reliable Tool-Calling Agent Loop on OpenRouter
OpenRouter · Sep 17, 2026
Two-Stage Personalized Gait Phase Estimation in Stroke Survivors During Exoskeleton-Assisted Walking: An Offline Feasibility Study
arXiv · Sep 15, 2026