AI text detectors misclassified over half of non-native English speakers' TOEFL essays as AI-generated in a 2023 Stanford study.

Is it true that AI text detectors misclassified over half of non-native English speakers' TOEFL essays as AI-generated in a 2023 Stanford study?

Unclear. Two sources partially support the claim: a 2026 systematic review reports that six experimental studies found AI detectors falsely labeled non-native English speakers' writing as AI-created in 50.2%–61.3% of cases, and a 2026 protocol cites a foundational 2023 study reporting a 61.3% mean false-positive rate on non-native English speakers' essays. Neither confirms that the study was conducted at Stanford or used TOEFL essays specifically.

Verdict: PARTIAL LEAN · 3 sources for, 0 against · Last checked 3 Oct 2026

The private library was unavailable for some searches in this check. Missing library results are not evidence against the claim.

JCY, N., C, K., & L, A. C. (2026). The impact of generative artificial intelligence tools on assessment equity for non-native English-speaking medical students: a systematic review.. doi.org. https://doi.org/10.1186/s12909-026-09303-7

Agha, A. & Anwar, E. (2026). Algorithmic Penalty for Non-Native Voices: A Three-Arm Diagnostic Accuracy Audit of Five Commercial AI Text Detectors Against Pre-LLM Scientific Literature — Protocol for the AUDIT-AI Study. doi.org. https://doi.org/10.21203/rs.3.rs-10164745/v1

W, L., M, Y., Y, M., E, W., & J, Z. (2023). GPT detectors are biased against non-native English writers.. doi.org. https://doi.org/10.1016/j.patter.2023.100779