In the age of artificial intelligence, this is the million-dollar question:
How can we prevent our students from cheating with AI tools?
AI tools such as Google Bard and ChatGPT are freely available for everyone to use (and abuse). Within seconds, these tools can write emails, create stories, and give lists of useful vocabulary, so it’s no wonder that students may be tempted to use AI to cheat on their language learning assignments.
In this blog post, I’ll explore some ways to ensure learners use AI responsibly.
1. Teach them to use it to help them learn
I have been using ChatGPT recently to help me learn Spanish. I have found it invaluable as I can ask it to give me translations, corrections, linguistic tips, example texts, explanations, and even regional differences or recipes.
I have never used AI to write an email, text, or homework assignment for me because I am aware that doing so would not help me learn. This is much like if a teacher gives students the answer without allowing them the opportunity to find it for themselves.
Teaching students to develop their AI skills will help them to use it to aid learning and will reduce the risk of misuse.
I’ve outlined some ways to develop learner autonomy and support language development using AI in this blog post: Developing Learner Autonomy with AI.
2. Make them aware of the consequences
I often found that when my students submitted work that was not their own, I could immediately tell.
Here’s why:
- Two (or more) students made exactly the same mistake (e.g., incorrect spelling).
- The work had no errors at all or minimal errors.
- They had used above-level language (e.g., a beginner using the word "submit" or the past perfect progressive).
If the suspicious work was word-processed, I’d often find myself copying and pasting it into a search engine to see what appeared. On occasion, I discovered that students had copied and pasted from websites.
Tools such as Turnitin and Grammarly can also detect plagiarism, but sadly can’t yet tell if a text has been generated by AI.
I always found that making students aware of the consequences of cheating and how teachers can identify it had a strong impact. Not only might they fail or be removed from the course if they get caught, but also their language learning will remain at the same level, which is a big motivator for many learners.
At the end of the day, learners who cheat are only cheating themselves.
3. Focus on the process
One of the best ways to prevent learners from cheating is by designing assignments that make it impossible!
Some ways to do this are:
- Incorporating unique contexts or case studies that are not readily available online
- Personalizing tasks so that students must write about themselves
- Providing real and local examples where students need to add their own thoughts and ideas
- Using a combination of assessment tasks (e.g., presentations, blogs, interviews)
At ElliiCon2023, Dr. Amin Davoodi advised teachers to use process writing strategies to ensure learners submit their own work. This involved:
- Writing a first draft of their work and submitting it to an AI tool for feedback
- Responding to the feedback and creating a second draft
- Submitting the final piece, drafts, and a copy of the AI prompts and responses
The students are then learning to use AI to support their learning and the teacher is also reassured that the work is original.
Here is a sketchnote summarizing the key points of Dr. Davoodi’s webinar:
It was absolutely fabulous, and I highly recommend you watch the full session to find out more (and why fridges can be scary)!
You can watch other AI-themed ElliiCon2023 sessions on the Ellii for Teachers YouTube channel. Subscribe so you don't miss a helpful video!
Related
Lessons
Blog Posts
What experience do you have of cheating and AI? What are your top tips for preventing it? We’d love to hear your ideas.
