AI incident ·
AI Detection Tools Allegedly Misidentify Neurodivergent and ESL Students' Work as AI-Generated in Academic Settings
In brief
An AI system built by Turnitin, GPTZero and 1 other and deployed by Universities, Colleges and 2 others allegedly harmed Students, Neurodivergent students and 6 others.
- Risk domain
- AI system safety, failures, and limitations
- Occurred
- Coverage
- 1 report
What happened
AI writing detection tools have reportedly continued to falsely flag genuine student work as AI-generated, disproportionately impacting ESL and neurodivergent students. Specific cases include Moira Olmsted, Ken Sahib, and Marley Stevens, who were penalized despite writing their work independently. Such tools reportedly exhibit biases, leading to academic penalties, probation, and strained teacher-student relationships.
Editor's notes
Reconstructing the timeline of events: (1) Sometime in 2023: Central Methodist University is reported to have used Turnitin to analyze assignments for AI usage. Moira Olmsted’s writing is flagged as AI-generated, leading to her receiving a zero and a warning. (2) Sometime in 2023: Ken Sahib, an ESL student at Berkeley College, is reported to have been penalized after AI detection tools flagged his assignment as AI-generated. (3) Sometime in late 2023 or early 2024: Marley Stevens is reported to have been placed on academic probation after Turnitin falsely identifies her work as AI-generated, though she purports to have only used Grammarly for minor edits. (4) October 18, 2024: Bloomberg publishes findings that leading AI detectors falsely flag 1%-2% of essays as AI-generated, with higher error rates for ESL students. (This date is set as the incident date for convenience.)
Laws that address this harm
Policy angle: Classified under AI system safety, failures, and limitations (Lack of capability or robustness) in the MIT AI Risk Repository taxonomy; 5 recorded instruments address this use case.
- India DPDP Act
- Law on Artificial Intelligence (2025)
- Law No. 132/2025 on artificial intelligence
- EU AI Act
- Texas Responsible AI Governance Act (TRAIGA)
Matched from the record's risk domain and country to the instruments recorded here. A reviewer can correct the match in the repository (data/external/incident_overrides.yaml).
News reports (1)
Titles link to the original publisher; report text is not reproduced here.
Who was involved
- Alleged deployer
- Universities Colleges Central Methodist University Berkeley College
- Alleged harmed party
- Students Neurodivergent students Moira Olmsted Marley Stevens Ken Sahib ESL students Epistemic integrity Educational communities
AI systems implicated
TurnitinGPTZeroCopyleaksAI-enabled decision support systemsAI detector technology
Classification (MIT AI Risk Repository taxonomy)
- Risk domain
- AI system safety, failures, and limitations
- Risk subdomain
- 7.3 Lack of capability or robustness
- Causal entity
- AI
- Intent
- Unintentional
- Timing
- Post-deployment
- Harm level
- —
- Sectors
- —
- Countries
- —
Risk entries describing this failure mode
Entries from the MIT AI Risk Repository coded to subdomain 7.3.
- Reliability issues
"Relying on general-purpose AI products that fail to fulfil their intended function can lead to harm. For example, general- purpose AI systems can make up facts (‘hallucination’), generate erroneous computer code, or pro...
- Type 2: Bigger than expected
Harm can result from AI that was not expected to have a large impact at all, such as a lab leak, a surprisingly addictive open-source product, or an unexpected repurposing of a research prototype.
- Type 3: Worse than expected
AI intended to have a large societal impact can turn out harmful by mistake, such as a popular product that creates problems and partially solves them only for its users.
- Ethics and Morality Issues
LMs need to pay more attention to universally accepted societal values at the level of ethics and morality, including the judgement of right and wrong, and its relationship with social norms and laws.
- Safe learning
"AGIs should avoid making fatal mistakes during the learning phase. Subproblems include safe exploration and distributional shift (DeepMind, OpenAI), and continual learning (Berkeley)."
- Malign belief distributions
"Christiano (2016) argues that the universal distribution M (Hutter, 2005; Solomonoff, 1964a,b, 1978) is malign. The argument is somewhat intricate, and is based on the idea that a hypothesis about the world often includ...
- Meta-cognition
"Agents that reason about their own computational resources and logically uncertain events can encounter strange paradoxes due to Godelian limitations (Fallenstein and Soares, 2015; Soares and Fallenstein, 2014, 2017) an...
- Technical and operational risks
"To date, technical limitations and vulnerabilities are present in most generative AI models in various contexts. Consequently, malicious users find it easier to breach an AI system’s safety and ethical guardrails to e...
Related incidents
Linked by editors or by text similarity in the source dataset.
Incidents in the same risk subdomain
- Purported AI Name-Reading System Reportedly Skipped and Misannounced Graduates at Arizona's Glendale Community College Commencement
- PocketOS Production Database Was Reportedly Deleted by Cursor AI Agent Running Claude Opus 4.6
- Baidu Apollo Go Robotaxis Stopped in Traffic During Reported System Failure in Wuhan, Stranding Some Passengers
- Purportedly AI-Enabled Targeting System Was Reportedly Implicated in Deadly U.S. Strike on Iranian Primary School
- Claude Code Agent Reportedly Deleted DataTalks.Club Production Infrastructure, Database, and Snapshots via Terraform
- Purportedly AI-Generated Sepsis Alert Reportedly Prompted Potentially Inappropriate IV Fluid Administration for a Dialysis Patient, Averted by Clinician Intervention
Other incidents involving Universities
Source record: incident #849 on the AI Incident Database · all 1 report