Skip to content

Ethical AI Use in the Classroom and a Freebie

Hands typing with text over the top Keeping the Human in the Loop!

Keeping the HUMAN in the Loop!

I had the opportunity to attend ISTELive 24 this past June. The common thread of almost all the sessions I attended was, “AI is here to stay and it is going to change education.” The conference was a whirlwind of excitement and buzz about AI in education and I was ready to jump in with both feet.

The morning after the conference ended, I sat at my gate waiting for my flight when I realized that I was going to need to start a conversation about AI use in the classroom with my high school students in the fall. I immediately knew how I wanted to approach it: through open conversation and as a co-learner with my students.

Nicole and I believe deeply that our AI use is only ethical when we keep our humanity at the center. With that said, we offer the acronym HUMAN to use with students as we teach them about the ethical use of AI.

Image from Wutthichai Charoenburi on Flickr

H- Halt. Is this something you want to use AI for? Is your outcome appropriate for the task?

When I started talking with my students about AI use in the classroom, they revealed that they had heard of ChatGPT, but none of them had tried it and they didn’t understand its capabilities.

I talked about the ways that I had been using it to support my work. I offered them ways that they might use it: as a brainstorming tool, as an idea generator, as a tutor, as a research assistant, and as a writing assistant (using it to help them outline or to get feedback on their writing). They were shocked – they had only thought about it in the context of “don’t use it because it’s cheating,” but had not yet thought about it as a tool. This led naturally to a conversation about thinking about ways that AI might streamline their workflow or be a thought partner for them and to think about how AI might be integrated into into their tasks.

I model this in class in front of them all the time. Last week we needed to write interview questions for elder interviews, so I posed the question, “Should we use ChatGPT to write our interview questions? Would that be an ethical use of AI?” Immediately, one student asked me, “It depends on what we are going to do with the questions.” When I explained that they would be used to conduct interviews and that the questions themselves were not part of the assessment, the students agreed that using AI in this way would be ethical and we proceeded.

U – Understand and check for bias.

Conversations about AI bias had not come up in class before my students experienced it. I had a group of female students who were using AI to generate ideas for titles for a podcast they were producing. After giving ChatGPT information that they wanted to incorporate into the titles, they got some interesting output. “Ms. Meyer, all the titles are coming back related to ‘Girl Power.’ It’s all ‘glitter, makeup, and clothes’ in our titles. What is going on?” When we checked their input, it was quickly evident that they had described themselves as “three female high school students,” which prompted the AI to generate titles that it “thought” related to girls. They were a bit miffed, but it opened up a spot for us to have a conversation about bias in AI. 

Fortunately, I knew enough about AI bias to begin the conversation, but we all quickly became co-learners as we started researching AI bias together. In 15 minutes or so we had found information, discussed it, and agreed that we’d share more examples of bias as it happened.

Not sure where to start when it comes to talking about AI bias? Common Sense Education has two excellent lessons for students in grades 6-12: Understanding AI Bias and How AI Bias Impacts Our Lives.

M – Make it yours! Does it sound like you? What can you add to this or do to it to make it yours?

One of the first ways I used AI with my students was to generate feedback on their writing. I gave my students some prompts to use that would help them get specific feedback on skills we were working on in class and they put in passages from their work to explore using AI for writing support. They were surprised by how accurate and detailed the feedback was. They were also surprised to find many inaccuracies and some poor feedback. We had a conversation about how to evaluate feedback and how they had the power as writers to choose to use feedback or not (from AI, from a peer, and, yes, even from me). We talked about how to prompt it to not rewrite our work but to offer suggestions. We explored how to use it to outline essays using their original ideas. And we agreed to disclose our AI use in our assignments when we had used it.

One of my students noted that using AI for feedback on his writing was helpful, especially when “it is 1 AM and your teacher isn’t available.”

A – Assess accuracy. Critically evaluate the accuracy of the AI’s output. Can the AI’s output be verified using other sources?

At the time of the writing of this blog, the free version of ChatGPT (which is what my students use) is still unreliable for factual information and I don’t use it with my students for anything that requires factual output. I have shown them how I have tried to use it to find resources for me and how it will make resources up (fake titles, authors, and descriptions of poems, articles, and YouTube videos).

We have tried Perplexity.ai as a research assistant, asking it to “find YouTube videos related to the theme of the destruction of the American Dream,” and that particular platform generates real resources and offers citations for any commentary, but that we still need to check any factual information from AI. I also introduced the idea of AI hallucinations – this is what it is called when AI makes up information and presents it as factual because of insufficient training. I will write an upcoming blog on how I taught this lesson using lateral reading – look for that in February!

If you’re interested in knowing more about how AI is trained, check out this lesson from Common Sense Education: How is AI Trained?

N – Next. What will you do with it next? Do you need to disclose that you used AI? Do you need to cite the AI?

When we first started talking about AI use in our classroom, we talked about plagiarism, an ongoing conversation in every ELA classroom. I had created a document for them that I had used AI to help me write. At the bottom of the document, I had a statement about my AI use, including the prompt that I used to generate the output they were seeing (because I want them to see good prompt engineering). As a co-learner, I asked them if they thought that statement helped me to avoid plagiarism. They agreed that it was definitely a step in the right direction and asked if there was a citation format they should follow. Together we read MLA Style Center’s page on citing generative AI in MLA format. We decided that when we used AI, to avoid plagiarism we would cite it in MLA format and include an annotation with the prompt they used and some reflection about their process of using AI.

My favorite question of the semester was when a student asked me, “If we accidentally plagiarize while using AI, will you tell us and help us figure out how to fix it?” I knew, at that moment, that we had reached my goal: we had become co-learners who were engaged in open conversation about our AI use. And that’s the best thing I could ask for. 

FREEBIE

Here is a tool for your classroom to help remind you and help you teach your students about ethical AI use.


Teachers Pay Teachers

Stay curious, stay innovative!

Two Maine Teachers Logo and words Kate

AI Transparency: Chat GPT was used to help develop this post.  Chat GPT 5%, Human 95%.  I use it to maximize the readability. Here is the prompt that I used:Prompt: “Read this blog post, DO NOT change the content.  Give me feedback on SEO* (Copy and Paste blog post)” *SEO is Search Engine Optimization.

1 thought on “Ethical AI Use in the Classroom and a Freebie”

Leave a Reply

Your email address will not be published. Required fields are marked *