In Spring 2026, 45.7% of students in a large CS2 course were flagged as LLM-assisted following structured manual review¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Manual review of the top-ranked students in Spring 2026 determined 267 of 584 exam-completing students (45.7%) had used LLM-assisted tools. → David Racovan 2026

Evidence¶

David Racovan 2026¶

David Racovan, Ajay Rawat, Christopher K. May, Jeffrey A. Turkstra. (2026). Argus: Academic Integrity in the Era of Generative AI. https://arxiv.org/abs/2609.36073 doi:10.48550/arxiv.2609.36073

q2 · i? · design · r2

Case study of the Spring 2026 offering, in which three course staff independently reviewed students ranked by H-score and reached consensus, giving extreme benefit of the doubt; "267 (45.7%) were flagged following manual review". No effect size is printed for prevalence.

"Of the 584 students who completed the first midterm exam in the Spring 2026 offering, 267 (45.7%) were flagged following manual review, and 317 were not."

Discussion¶

Related Claims¶

• Flagged Spring 2026 students scored lower on all three exams than non-flagged peers, with performance declining as flagged assignments increased — related
• MOSS detected far fewer suspicious cases in Spring 2026 than in prior offerings, as LLM-assisted submissions lack similarity signals — related
• Average H-scores rose from near-zero in pre-LLM offerings to 11.9 in Spring 2026, with maximum scores reaching 81 — related