Preventing AI-Based Cheating

AI tools like ChatGPT, Claude, and Gemini have transformed how students access and generate information. While these technologies offer significant educational benefits, they also introduce new challenges—particularly around academic integrity.

With just a few prompts, students can generate entire essays, solve assessment tasks, or summarise readings—without fully engaging in the learning process. For educators and institutions, this raises a critical question: How can we protect the value of assessment in the age of AI?

This blog outlines practical strategies to reduce AI-assisted cheating in both online and in-person assessments—while still embracing AI’s potential as a learning support tool.

1. Design Smarter Assessments

The first line of defence is the assessment design itself. Tasks that require original thought, personal experience, or real-time engagement are much harder to outsource to AI.

Authentic Tasks

Ask students to apply learning to real-world situations, local case studies, or personal reflections. These tasks are harder to generate generically and easier to assess for genuine engagement.

Oral Defences (Viva Voce)

After submitting an assignment, students can be asked to explain their work in a short verbal follow-up—either in person or via recorded video.

Progress-Based Projects

Break large assessments into milestones or drafts. This shows student development over time and discourages last-minute AI-generated work.

In-Class Writing Tasks

Occasionally revert to supervised, handwritten assessments. These provide useful baselines to compare against digital submissions.

2. Secure the Assessment Environment

If you're using an LMS or online testing tool, there are several steps you can take to make it more secure.

Disable Copy-Paste in the LMS

Prevent students from pasting AI-generated answers into text boxes. Most LMS platforms allow for disabling right-click and copy/paste in exam mode.

Use Lockdown Browsers

Browser lockdown tools restrict students from opening new tabs, switching windows, or accessing AI tools during online exams.

Randomise Questions

Use question banks that rotate and randomise content for each student. This makes AI-generated answer sharing ineffective.

Time-Limited and Adaptive Tests

Shorten timeframes and use adaptive question delivery to reduce the chance of students consulting external tools.

In-House, Supervised Testing

For high-stakes exams, bring students onsite to complete assessments under staff supervision. This ensures a controlled environment and full academic integrity.

3. Leverage AI Detection Tools

Even with strong design and security, some students will still attempt to use AI. Here’s how you can detect it.

Use AI Plagiarism Detection

Platforms like Turnitin (AI Writing Indicator), GPTZero, and Originality.ai are built to detect AI-generated text and flag suspicious patterns.

Collect Baseline Writing Samples

Have students submit early writing samples in class. Use these to compare against future assignments and look for sudden changes in tone or structure.

4. Build a Culture of Academic Integrity

AI is here to stay. Rather than fight it completely, institutions should teach students how to use it ethically and understand its boundaries.

Update Your Policies

Clearly define what constitutes AI misuse. Is using AI to brainstorm allowed? What about editing? Make it transparent.

Educate Students

Teach digital literacy and ethical use of AI. Show how it can be used responsibly—for outlining, summarising, or reviewing work—but not to generate full submissions.

Set Clear Expectations

Make your rules explicit in assignment briefs. Include a statement requiring students to confirm whether AI tools were used and how.

Model Ethical Use

Encourage teachers to demonstrate responsible AI use—such as using it for lesson planning or feedback examples—so students see it as a tool, not a shortcut.

Final Thoughts: It’s About Balance, Not Bans

Banning AI outright is neither realistic nor educational. Instead, institutions must find a balance—using smart design, secure platforms, clear policies, and proactive education to maintain assessment integrity while preparing students for the AI-enhanced world ahead.

Practical Next Steps for Educators:

• Review and redesign assessments for authenticity

• Enable secure testing settings in your LMS

• Use AI detectors with caution and human judgment

• Talk openly with students about expectations

• Encourage ethical, supported use of AI tools