Cheaters Never Win, But Are They Really Cheating? Rethinking Integrity in the AI Classroom

When Katrina Overby, an assistant professor of communication at the Rochester Institute of Technology, discusses academic integrity, she moves the focus away from detectors and toward a fundamental reframe: cheating as "outsourcing thinking".

This shift, from a violation of rules to a form of self-sabotage, was the central theme of our recent webinar. Educators across disciplines gathered to address an increasingly urgent question: what does academic integrity mean in an AI-enabled classroom?

Why do students reach for shortcuts?

The impulse to use AI tools often stems from a lack of preparation or confidence rather than a desire to be dishonest. Sarah Lahman, a professor of biology at the University of Mount Olive, observes that students who have been rewarded primarily for memorisation often panic when faced with high-stakes application tasks.

Common factors that converge to encourage shortcutting include:

• A Lack of Agency: Students may not feel equipped to meet the creative demands of upper-level courses.
• Time Management Struggles: When students feel overwhelmed, they reach for the nearest available tool to complete the task.
• Grade Anxiety: The fear that grades will suffer in unfamiliar assessment formats can drive students toward AI.

How can educators design assignments that make learning visible?

One of the most effective ways to support integrity is through authentic assessment design that prioritises the learning process over the final product. The webinar highlighted two practical models:

1. The Cumulative "Alien Planet" Project

Sarah Lahman’s ecology students design a complete ecosystem for a fictional planet with unique environmental conditions invented fresh each semester. This design "holds up" against AI because:

• It requires layered context: The specific gravitational and atmospheric conditions do not exist in standard AI training datasets.
• It uses AI as a coach: Students use AI for image generation under supervision, turning the exercise into a lesson in precision and asking better questions.
• It offers multimodal choice: Students present their findings through posters, videos, or even Minecraft environments, which reduces the anxiety of a single high-stakes format.

2. Collaborative Process Modeling

In her public speaking courses, Katrina Overby builds speeches collaboratively in class. Groups debate the strength of various sources and watch as the educator constructs an argument in real time. This process demystifies how research becomes structured, helping students produce better original work in subsequent assignments.

What uniquely human skills remain essential?

While AI can replicate a specific tone, it cannot embody the lived experience or credibility that students bring to their work. The panelists identified several core skills that educators can explicitly foster:

• Autonomy and Agency: The ability to make independent choices.
• Information Literacy: Precision in language and the evaluation of scientific claims.
• Metacognition: Reflecting on one's own thinking and learning process.

How can educators co-create AI rules with students?

Ambiguity regarding AI expectations often leads to confusion or inconsistent enforcement. Leslie Allen Essex, Macmillan’s Senior Marketing Manager for Technology, demonstrated a replicable activity where educators and students co-create an Integrity Charter.

By discussing specific scenarios, such as using AI to brainstorm a thesis or translate a draft, educators can help students define what counts as a meaningful learning commitment in their specific course. Students who help write the rules are often more invested in following them.

Moving forward with confidence

Rethinking integrity for the AI era does not require a total course overhaul. This tension is not entirely new; educators have always managed challenges to assessment, from cheat sheets in exam halls to formula memorisation.

The instructors doing this work best acknowledge they are still learning alongside their students. Even mid-semester, educators can take baby steps by starting a conversation with students about which skills the course is building and why they matter for their futures.

Further Reading

Assessment in the Age of AI: What Higher Education Instructors Actually Need to Change in 2026

The Detective Fatigue is Real: Why the Best AI-Proof Courseware Focuses on Support, Not Surveillance

Cheaters Never Win: From Cheat-Proof to Learning-Rich Assessment Design Webinar Recording

FAQs

How can I make my assignments AI-resistant?

Incorporate "dirty data" or experimental anomalies that require human interpretation. Asking students to explain why their real-world results differed from a theoretical model is much harder for an AI to replicate convincingly.

Does AI detection software still work in 2026?

While detection tools exist, they are often a step behind evolving LLMs. Most experts recommend focusing on pedagogical intervention and authentic assessment rather than relying solely on technical detection.

How does Macmillan Learning support academic integrity?

Macmillan Learning focuses on intrinsic motivation. Tools like Achieve provide formative assessments and metacognition features that help students feel prepared, reducing the stress-induced urge to use AI shortcuts.

What is Achieve?

Achieve delivers research-backed personalised learning, real-time feedback, and accessibility features—all integrated into your LMS to keep students engaged without increasing your workload.