Recently, the increasing popularity of AI technology through ChatGPT and similar generative AI apps has led our school to create a policy regarding the use of such tools in educational settings. In the recently updated 2023-2024 student handbook, it now reads, “Students at Belmont Hill may not use AI generative tools like ChatGPT for school work, homework, or assessments unless their teacher has given them explicit instruction to do so. Using work generated in part or in whole by any AI tool will be considered a violation of Belmont Hill’s honor code and will result in a disciplinary response.”
After reading this short policy to a couple of Form IV students, both agreed that the policy is vague. To start, the policy prohibits the use of work generated by AI tools in homework, but doesn’t clarify all uses of artificial intelligence. For example, does it mean that students cannot use ChatGPT to proofread their work? Is there a difference between using a tool like Grammarly and ChatGPT? The word “generated” in this context implies ONLY that individuals are not allowed to copy and paste ChatGPT responses into their work. What about to draft ideas, or to gather basic background information such as one can find on Wikipedia? These are questions that many students and faculty have and continue to consider. While neither of the students interviewed condone cheating, they assert that there are instances where ChatGPT could be used without violating the Belmont Hill honor code.
The rise of generative AI technology in the near future is inevitable and will likely be applied everywhere very soon. As Belmont Hill aims to preserve the strict academic integrity that has been a hallmark of the school since 1923, it finds itself at a crossroads. There are two paths our school can take. The first is that the school can ban the use of ChatGPT and generative AI completely. Block the website from computers and change the policy to explicitly say that ChatGPT (or other similar systems) can’t be used at all. However, this policy won’t achieve much as most students have access to VPNs on their laptops. The second option is to invite students and faculty to develop more clear protocols about constitutes “fair use” vs. cheating when it comes to use of generative AI technologies. Increased participation in the policy development process will increase buy-in and improve the policy overall. That being said, as the technology is evolving so quickly, a school policy can’t possibly cover every possible use of ChatGPT. Inevitably, there will be specific disciplinary cases that will need to be handled on a case by case basis.
If the administration takes the proper steps, generative AI can be harnessed to help students study, instead of being viewed as an impediment to learning. Many well-known online resources like Quizlet and Khan Academy have already implemented generative AI as a tool. There are many other examples where AI generative technology is used as a learning tool for students. The school should allow students to use these sorts of tools instead of banning AI generative tools completely.
The growing impact of AI technologies can’t be ignored. However, the current policy’s vagueness leaves room for interpretation and raises questions about the boundaries of AI tool usage in academic settings. The question remains: can Belmont Hill find a balance between maintaining academic integrity and accepting the potential benefits of AI?
