Since the introduction of ChatGPT, OpenAI’s language model has had a profound impact on numerous industries, including schools and educational institutions. As a result, questions regarding how educators should approach discussing this topic with students, how teachers and parents can identify AI-edited text, and what measures can be taken to prevent cheating facilitated by ChatGPT have become increasingly vital.
With every homework assignment and text produced by students, it has become imperative for schools and teachers to acknowledge that ChatGPT may have been involved. However, as a teacher or parent, it can be difficult to determine whether AI has been used in the final result. It is also worth considering the significance of this knowledge and whether the introduction of ChatGPT should prompt us to think more deeply about the need to explore alternative exam formats.
A frequently asked question is why students have access to AI tools like ChatGPT if registration is only allowed from the age of 18. In addition to faking your age, there are other ways to access open AI applications. For example, Snapchat has been providing access to ChatGPT for a few months now. Every Snapchat user now has “My AI” as their first contact and can access ChatGPT in a normal chat, ask questions, get help with homework, or even send pictures. This means that even students under the age of 18 have access to AI tools like ChatGPT and can use them in a variety of ways.
Many teachers are now rightly wondering how they can technically verify whether the students have done their own homework or have involved ChatGPT in some form.
The short answer: they can’t.
The long answer: You know your students best, and you will quickly notice if the quality of the homework suddenly improves, or if the content cannot be adequately reproduced in oral follow-up questions. Parents will also notice these changes when they look at their children’s homework.
However, many teachers and parents would like to have tools and technical aids that allow them to enter a text and easily check whether it was written by the student or directly from ChatGPT.
These types of tools exist, but none of them can tell you with 100% certainty whether a text is AI-generated or not. For example, in a handout for teachers provided by OpenAI, they point out that using AI detectors is not effective, as ChatGPT ultimately has no knowledge of whether a text is AI-generated or not. On the other hand, banning ChatGPT with technical verification by an AI detector is also not effective in the long run. Students will have access to these tools, therefore schools and parents cannot simply turn a blind eye to ChatGPT.
Nevertheless, the question remains: how can teachers and parents determine whether or not students are using AI? The simplest solution is to allow the use of technology such as ChatGPT during certain phases of the lesson or for certain assignments, while establishing clear guidelines for its use in collaboration with students. After all, ChatGPT is a technological tool, much like a calculator. Just as students must first learn numbers and basic arithmetic, understand basic mathematical principles, and understand how a calculator works before they are allowed to use one, a similar approach can be taken with ChatGPT. By aligning the introduction of ChatGPT with appropriate grade levels, the tasks can be adjusted accordingly.
The same goes for ChatGPT: students require a proper introduction and explanation before using the tool. They need to develop a basic understanding of how it works and how it generates text. Through transparent discussions and explanations, and by clearly outlining the circumstances and guidelines for using AI tools, as well as how their use will be evaluated, teachers and parents can eliminate the need to differentiate between AI-generated text and student-generated text. Furthermore, citing ChatGPT as the source makes it easy to see when it has been used. Through open communication and clear guidelines, teachers and parents can confidently identify the use of ChatGPT in student work.
While it’s true that AI tools can potentially be used in cases where they shouldn’t be, parents’ and teachers’ fears of “AI cheating” can be mitigated by promoting transparency and establishing explicit grading guidelines for work involving AI tools. By exploring AI tools together, setting clear rules, and engaging in meaningful dialogue with students, educators can instill a reflective approach to the use of AI tools. This proactive approach eliminates the need for ineffective AI detectors. Ultimately, through open communication and guidance, students can develop responsible and ethical practices when using AI tools.
In November, ChatGPT will be one year old, and it has already found its way into classrooms and children’s rooms, whether through Snapchat, Roblox, or simply by creating an account on OpenAI. This also means that schools and parents cannot ignore artificial intelligence – the topic of artificial intelligence must be addressed, discussed, and the critical yet reflective use of AI tools must be explored.
Innovation meets well-being: schools in Denmark
