AI evidence extraction

Cheating in the second year of generative AI chatbots: a follow-up study on high school student cheating behaviors

Ruishi Chen, Victor R. Lee, Annie Camey Kuo, Denise Clark Pope, Sarah Miles · Educational Technology Research and Development · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

7/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

12.63

FWCI

Top 10%

Impact

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1007/s11423-026-10587-1

Methodology & findings

Study design

Comprehensive survey of students across six schools (N = 4,354) in the United States, investigating self-reported cheating behaviors, patterns of AI use across different school-related tasks, and student perspectives on appropriate AI use in academic settings..

Sample

N = 4354, 2 groups

Main result

The study found that "overall cheating rates remain stable at 72.06%, consistent with historical baselines and prior studies, suggesting that AI availability has not changed overall cheating prevalence in high school." Additionally, "more students reported using AI chatbots for support tasks like concept explanation and idea generation," and students "still strongly supported using AI for conceptual understanding and brainstorming, and they maintained clear boundaries against using it for completing entire assignments."

Reports effect sizes.

Research paradigm

Positivist/empiricist

Author conclusions

The authors conclude that "while AI's prevalence has not altered the patterns of academic integrity at schools, students' evolving perspectives on appropriate AI use provide valuable insights for schools and administrators integrating AI into traditional school settings."

Risk of bias

Self-reported cheating behavior (social desirability bias); Selection bias (only surveyed students at six schools, may not represent all high schools); Cross-sectional design (cannot establish causality); Self-reported cheating behaviors (social desirability bias); Survey-based measurement without objective verification of cheating claims; Potential selection bias in school sampling (only six schools); Cross-sectional design cannot establish causality; No details provided on response rates or non-response bias; Self-reported data on cheating behaviors (social desirability bias); Survey-based methodology (response bias); Selection of six schools may not be nationally representative; Timing of study (one and a half years after ChatGPT release) reflects a specific snapshot in time

Open questions raised

The abstract indicates that students' evolving perspectives on appropriate AI use provide insights for future integration of AI into school settings, though specific future research directions are not detailed in the abstract.

Data: not_statedCode: not_statedExtracted from: pdfAgreement 61%

Explore related topics

Responsible AI Use Research Integrity Survey