What does cheating with AI really mean?

In May 2026, the largest (to date) study of undergraduate AI usage offered one of the first systematic looks into the specifics of how and why students use AI. To the researcher’s estimate, 9% of students who use GenAI have “submitted AI-generated content despite knowing it may not be allowed.” Universities, including NYU, are wrestling with the potential for AI to be used to commit academic dishonesty.

The Gray Zone of Student Use

Generating a research paper wholesale or copy-pasting sections of text from an AI chatbot is a clear cass of academic dishonesty. However not all cases are so cut and dry. The everyday reality for students may be more nuanced:

• Grammar & Style: Does running a draft through an AI proofreader count as an integrity violation? 

• Feedback: Is asking a chatbot to look for gaps in an argument or for suggestions for syntax acceptable if the student rewrites their work themselves ? 

• Ambient AI: In a classroom with a strict “No AI” policy, does reading the AI summary at the top of a Google search count as cheating? 

The myriad in-between, gray areas combined with the pervasiveness of AI tools have many professors walking a tightrope. Students may also be guessing about exactly what is or isn’t allowed and why. In a complex academic and technological environment itself, ambiguity may lead students to default to over-reliance or to avoid tools altogether. 

Shifting to Participatory Communication

Opening up clear and participatory communication can alleviate some of the pressure. Wrestling with the nuance and implications of AI use in the classroom may also cause some of the productive struggle that leads to deep learning.  

When assigning work, consider a few ways to get that conversation started:

• Co-create Norms: Ask the class what kind of AI usage they think is fair or unfair for the assignment. 

• Define Scope: Clarify explicitly exactly when AI is okay or not okay: research, brainstorming, designing, editing, drafting?

• Address Tools: Define which platforms are recommended and NYU-approved, and how students should cite or document usage. 

• Discuss Learning Trade-offs: Provide clear rationales for your AI policy and explain why certain constraints exist. Ask students to weigh in. 

• Demonstrate Positive Uses: Show students acceptable ways to use AI on an assignment and discuss strategies with the class. 

Transparent Pathways

Inviting students into open dialogue may prevent the ambiguity and misunderstandings that can lead to academic integrity issues. NYU offers guidance on adapting coursework for an AI era, including recommendations for class policies and syllabus language to help establish clear boundaries without turning to an over-reliance on surveillance. 

Have you experimented with dialoguing with your class on AI guidelines? Contact the Nexus team at nexus@nyu.edu for guidance and support with adapting AI-informed assignments and defining your course policies.