Language Learning with VR and AI - 2026 

Project Overview:

The iTLAB and Anna E. Venetsanos, Clinical Associate Professor of Modern Greek, began a successful initiative in 2025 to enhance Elementary and Intermediate Modern Greek courses. We have continued in 2026, with students using the Meta Quest 3 XR headset, Mondly, and an AI-powered simulation built with Wonda Spaces.

The original 2025 goals persist:

• Technology-supported, student-centered approach to language learning.

• Enabling students to practice speaking, review mistakes, and learn in a low-risk environment.

• Building student confidence by allowing practice without fear of errors.

Bringing the Experience to Life:

The session structure is also the same as in 2025, with students attending guided XR sessions at the iTLAB. Experiential Activities include:

• Structured scenarios (e.g., taxi, travel conversations).

• Freeform conversation with an AI avatar around vocabulary, grammar, and language elements in accordance with the CEFR* and ACTFL**

*Common European Framework of  Reference for Languages

**American Council on the Teaching of Foreign Languages

A primary difference is that in 2025, an in-house application was used for the AI component. Although customizable, software support proved unsustainable. In 2026, we switched to Wonda Spaces, a vendor platform used by NYU.

Refining the Models:

The Wonda platform allows us to swiftly customize each session's AI prompt, allowing Professor Venetsanos to tailor material based on student proficiency, even on short notice. The Mondly experience of structured scenarios remains largely the same (see details in 2025 case study), and continues to satisfy a Contextual Repetition model.

Original challenges and mitigations persist (see details in 2025 case study). New 2026 challenges include optimizing the AI prompts to provide the best student experience. The mitigation is to revise based on straightforward student feedback. 

The mechanics of revision require a fair amount of trial and error testing of the prompt.

Outcomes - Positive Feedback:

• “I was able to have a realistic speaking experience, and it challenged me to really dive deep into what I have learned so far this semester.” 

• “Practiced conversation skills in Greek like it was real life.”

• “I had a better experience this time around, and she spoke better.”

• “The appointment thing was different. Really liked it, even if the answer was not accepted. I was put on the spot..if you redo it it’s good!!”

• “I reałly enjoyed this last VR experience. Athena has become much easier to talk to. I can speak to her in English, and she will respond. She is able to repeat phrases, adapt to my level, and continue a conversation.” 

Outcomes - What could be improved?:

• “Adjust the talk speed and maybe add the speech bubble so we can read what they’re saying as well, since she talks so fast.” 

• “Overall, it was good, but she would either tell me to slow down or repeat.”

• “For whatever she said, I would like it to come up on the screen and then disappear, like subtitles.”

• “I enjoy Mondly very much; only thing was different was that I was not looking at suggestion box and would say something not accepted what I would say wrong and try again.” 

• “First time in mondly—it didn't understand me at all, but i had to give simple answers to understand but it felt like it was a2 level.”

What We learned:

Although scaling virtual reality at universities remains challenging, scaling AI roleplay using platforms like Wonda Spaces is achievable. And students appreciate and value that virtual reality provides a different learning experience than traditional methods, even if it is a small-scale resource.

Acknowledgments:

This work builds on the curiosity and enthusiasm of the students who engaged with it across both years. Their willingness to speak in a new language, to an AI, in a headset, and tell us honestly what worked and what didn't is what made iteration possible. The refinements visible in this 2026 study are, in a real sense, their work as much as ours.