Research

AI Cheating in Online Exams: Primary-Source Research and Statistics

Primary-source research on AI-assisted cheating, assessment fraud, and the tool ecosystems that have made detection architecturally impossible.

Primary-source reports

Quarterly

Threat index cadence

Datasets

Annually updated primary-source datasets covering cheating rates, tool ecosystems, and financial impact.

Dataset

Updated March 2026

AI Cheating Statistics 2026

A comprehensive, annually updated dataset covering AI-assisted cheating rates, fraud trends, tool ecosystems, regional breakdowns, behavioral detection signals, and the financial cost to employers.

35% of proctored assessments showed AI-assisted fraud attempts in 2025, up +119% year over year (CodeSignal); 88% of online exams now face active AI cheating risk (Talview).

Read the report →

Indexes

Quarterly-refreshed rankings tracking specific threat tools and defense effectiveness over time.

Quarterly Index

Updated Q2 2026

Threat Index: Q2 2026

Quarterly tracking of invisible AI overlays, remote-access tools, proxy services, and the defenses that attempt to stop them. Dataset published under schema.org/Dataset.

6 of 7 tracked defense categories have at least one documented, working bypass; only one has none for enrolled-device, internet-dependent cheating.

Read the report →