AI writing and how to manage it in the classroom was top of mind for the education professionals who traveled to Indianapolis for the International Center for Academic Integrity’s (ICAI) annual conference March 9–12.
Educators, administrators, researchers, and integrity officers who attended the event all agree on one thing: The challenges that AI writing tools present are uniting the education community around the need to identify effective measures to manage them. However, what these measures should look like is up for debate.
First Consider the “Why”
Before delving into potential changes to educational policy around AI writing in the classroom, it’s important to first answer the following questions:
- Why are writing and original thought needed?
- What purposes do writing and original thinking serve?
Only after serious consideration of the above questions can an institution decide how their policy should respond to AI writing.
For example, educators assign writing prompts to their students to measure their grasp of the topic and to sharpen students’ critical thinking skills. In this case, limiting the use of or banning AI writing tools altogether might be a solution.
However, there are examples where using AI writing tools might be helpful. For researchers, AI writing tools could allow them to concentrate on their logical train of thought rather than focusing on the technical writing process. Furthermore, students with dyslexia, those who speak English as a second language, or students with vision impairments could benefit from using those tools to help them focus on work content.
Over the three-day conference, these questions were discussed regularly. It was clear to many that changes are underway, and that no matter how they are answered, the educational community needs to rethink what assessment of knowledge means and looks like in the classroom.
What to Expect Next
During an AI writing session, ICAI attendees reviewed two text samples and were asked to identify which one was written by AI. While it was easy for attendees to discern which sample was AI-generated content, it was difficult for them to explain why they knew the sample was written by AI. As with contract cheating, it is important for educators to have the data to back up their intuition to remove bias and uphold integrity.
For writing evaluation tools like Turnitin’s solutions, it is already possible to detect AI generated text with an accuracy of up to 97%
During the conference, it was clear that attendees were ready to learn more about AI technology, determine strategies to manage AI writing, and identify opportunities to leverage its capabilities to better student learning.
For more information on Turnitin’s AI writing detection efforts, visit the AI writing detection landing page, and check back regularly for updates.
A version of this post originally appeared on the Turnitin Blog.
Sponsored Content Disclaimer:
Sponsored Content in Partnership With NASSP
NASSP allows select groups to share information and thought leadership with our program audiences.
