Nicola Work, OWLA President-Elect
Associate Professor of French, University of Dayton 

“Turns out the conversation partner my students would actually use was available 24/7 and never got tired of repeating the same question.”

AI is everywhere—and it’s here to stay. Students use it. Teachers use it. Administrators use it. Instead of forbidding it or pretending it doesn’t exist, we can learn to work with it and explore what it can actually do for us as language teachers—and for our students.

For years, I’ve wanted to give my students more opportunities for conversation practice. Of course, we can (and do) practice in class, but class time is limited. I wanted to offer additional practice. The question was: how?

I tried conversation hours (think office hours, but in French). Attendance? Spotty at best. I assigned partner recordings through our online workbook, but students struggled to coordinate schedules—and, frankly, they hated it. What I needed was a solution that was low-anxiety, easy to use, and always available.

Enter AI.

At a language-teaching conference, I attended a presentation where a colleague demonstrated SchoolAI. He showed how students could interact with an AI chatbot in a controlled, safe environment—either through writing or speaking. The examples included dialogue practice, picture descriptions, reading comprehension activities, and more. Naturally, I tested it out using my Spanish, and I was intrigued.

I tried SchoolAI with my students, and it works beautifully. You can create practice conversations using a vocabulary bank or sample questions. You can decide how many questions are asked, in what order, how the AI responds, how long the conversation lasts, what errors are corrected, and what kind of feedback is provided. Then—boom—you have a conversation partner ready to go.

What makes this especially powerful is that the conversation is different every time. Students can practice as often as they want, the experience is low-anxiety, feedback is immediate, and the platform does not collect student data (students only submit their first name). The possibilities are endless.

This semester (and last), I’ve been doing something similar using Google Gemini. Since we are a Google school, it integrates seamlessly into Canvas. I learned how to create custom GEMs that function much like what I built in SchoolAI.

For example, I created a custom chatbot to practice greetings in French. Behind the scenes, I gave the bot very specific instructions: what questions to ask, how many, what level of French to use, whether corrections should be made, what to do if a student answers in English, and more. Once set up, students could practice by typing on their computers—or, if they used the Gemini app, by speaking directly to the chatbot.

This activity can easily be replicated for other practice scenarios: ordering food, talking about likes and dislikes, describing one’s family, Q&A sessions about class topics, or discussions related to a film or reading.

I also used Google Gemini to create reading comprehension activities. Again, behind the scenes, I provided the target vocabulary, specified the text length, determined the number of questions, and set the language level. The result? Students receive a new, customized reading comprehension activity every time they log in—unlimited practice with immediate feedback.

Once I got started, I didn’t stop there. I  created a basic grammar tutor, a verb conjugation practice partner, and a chatbot that generates an image and asks students questions about it.

In the short time that I’ve been using these tools, students have been intrigued—and, more importantly, they’ve enjoyed practicing with them.

If you’re interested in learning more about AI and how it can support language teaching, join us at OWLA’s Spring Conference at Ohio Dominican University in March. We’ll be offering pre-conference workshops on AI, as well as conference sessions focused on practical classroom applications. We would love to see you there!