You ask your students to write a paragraph. Inevitably at least some students will try to use AI to write it for them, no matter how many times you warn them that you will know.

So, with AI getting smarter, how can we know?

Artificial Intelligence (AI) tools have become increasingly sophisticated, making it easier for students to generate written content with minimal effort. While these tools can support learning when used responsibly, they also pose challenges for educators striving to ensure authentic student work. So, how can you identify AI-generated writing—and guide students toward ethical use?

Where to Start

Before diving into guidelines and tools, it’s essential to start with open discussions about AI. Talk with students about when and why AI might be appropriate—for example, brainstorming ideas or improving clarity—versus when it undermines learning objectives. These conversations set expectations and help students understand that AI is a tool, not a shortcut.

I also make it clear that my role is not just to detect AI use but to examine writing for originality and authentic thought. I explain that I look for evidence of their unique voice, critical thinking, and incremental progress in their drafts. This transparency builds trust and reinforces the importance of academic integrity.

Tips to spotting AI Writing

Run writing through an AI detector.

 AI writing is getting more natural every day and, although I am pretty good at spotting it, I’m not great anymore. There are a bunch of free AI detectors that have done a great job for me. I always follow up with tip 2 and 3, but I start with the AI detector.

How I tested the AI detectors. There were three parts to testing.

1. I took some AI created writing and put it into Note GBT.  
2. I put in some of my own writing. 
3. I put in half AI writing and half my own writing. I then changed the order of the writing, my half before AI’s half, to see if it made a difference.

AI Detector Comparison Chart

AI Detector	Test 1: AI Text	Test 2: Human Text	Test 3: Mixed Text	Observations
Note GBT	100% AI	100% Human	100% AI (order change made no difference)	Consistent but not flawless—mixed text still flagged as 100% AI
Eduwriter	100% AI	100% Human	93% AI (human first), 100% AI (AI first)	Similar performance to Note GBT, slight variation when human text appeared first
Skribbr	100% AI	100% Human	100% Human (human first), 50% AI (AI first)	Showed the most inconsistency, sometimes labeling mixed text as fully human

Key Insight: AI detectors are helpful but not infallible. Use them as an initial check, not a final verdict.

Given the inaccuracies of the AI detectors, I will check the student’s website usage next. 

 My district uses Goguardian, so I can look at a student’s website history when they use their Chromebook. If I see an AI writer in their history, I now have enough information to talk to them.

If I still suspect AI usage

Have a conversation with students about the piece of writing. I will ask students to show me their document’s history. This allows me to see edits they have made as well as the writing process. I use this step regardless of what steps 1 and 2 have told me, so that I can understand the student’s process. This also gives students a chance to be honest with me and fix their mistakes. 

Pro Tip: This step is essential regardless of previous findings. It promotes transparency and gives students an opportunity to correct mistakes.

Teach Students to Use AI Responsibly

Once students understand that you can detect AI-generated work, shift the conversation toward responsible use. AI is not inherently bad—it can be a powerful learning tool when applied ethically. Here are strategies to guide students:

Model Appropriate Use

• Demonstrate how AI can assist with brainstorming ideas, improving clarity, or checking grammar, rather than replacing original thought.
• Show examples of AI-generated drafts and discuss their limitations (lack of personal voice, creativity, or nuanced understanding).

Set Clear Guidelines

• Define what constitutes acceptable AI use in your classroom (e.g., idea generation vs. full-text submission).
• Incorporate these guidelines into your syllabus and discuss them openly.

Integrate AI Literacy

• Teach students to critically evaluate AI outputs for accuracy and bias.
• Encourage them to cite AI assistance when used, just as they would any other source.

Promote Reflection

• Ask students to submit a short reflection explaining how they used AI in their process.
• This fosters accountability and metacognition.

“Our goal isn’t to punish students for using AI—it’s to teach them how to think critically and use technology as a tool, not a crutch.”

Ultimately, my goal isn’t to punish them. It is to teach them to use AI responsibly and to think for themselves. By combining detection strategies with education on ethical AI use, we empower students to navigate a world where technology is ubiquitous. Integrity and digital literacy go hand in hand.

Want to learn more about integrating AI responsibly in education?

Attend many of the educator-led sessions at the NCCE 2026 Conference and explore NCCE’s professional learning opportunities at https://ncce.org. Equip yourself with strategies to guide students in ethical technology use and join a community of forward-thinking educators.

---

Kate Metz, Independent School District 1 Science Teacher

10 years in Education, BS Biology MSU Denver