PromptRiskDBThreat intelligence atlas

Cheating/plagiarism

AI Risk

"Cheating/plagiarism - Use of another person’s or group’s words or ideas without consent and/or acknowledgement."

Overview

A source-backed snapshot of this AI risk.

Techniques0Attack methods connected to this risk.
Mitigations0Defenses that may help with related attacks.
Records2Source records unified into this concept.

Risk profile

How the MIT AI Risk Repository categorizes this risk.

Domain4. Malicious Actors & Misuse
Subdomain4.3 > Fraud, scams, and targeted manipulation
Entity3 - Other; 1 - Human
Intent3 - Other; 1 - Intentional
Timing3 - Other; 2 - Post-deployment
CategorySocietal and Cultural
SubcategoryCheating/plagiarism

Merged risk records

Source records unified into this canonical risk concept.

2 recordsView all →

MITRISK-Abercrombie2024-58.07.02 - Cheating/plagiarism

"Cheating/plagiarism - Use of another person’s or group’s words or ideas without consent and/or acknowledgement."

Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulationSourceA Collaborative, Human-Centred Taxonomy of AI, Algorithmic, and Automation HarmsYear2024

MITRISK-Li2025-66.04.05 - Cheating / plagiarism

"Use of generative AI in an academic setting to either cheat or plagiarize"

Domain4. Malicious Actors & MisuseSubdomain4.3 > Fraud, scams, and targeted manipulationSourceA Closer Look at the Existing Risks of Generative AI: Mapping the Who, What, and How of Real-World IncidentsYear2025

Related techniques

Attack methods connected to this risk.

View all related techniques →
No linked attack methods. No AI attack method is connected to this risk in the current data.

Mitigations

Defenses that may help with related attacks.

View all mitigations →
No propagated mitigations. No defense is available through the connected attack methods.

Source evidence

Research source for this risk, when available.

View all sources →