‘Generate a visual depiction of the party scene in The Great Gatsby.’
‘Explain how latitudinal gradation occurs using the AP Biology rubric.’
As artificial intelligence becomes more publicly accessible, users, especially students, are changing the way they interact with the internet for information. Instead of typing questions, queries now often take the form of direct commands. This year, Hill implemented an AI assessment scale to set students and teachers on equal footing when navigating how AI can and should be used in academic settings.
Hill first introduced a school-wide AI policy in the spring of 2023, when faculty recognized that AI tools would inevitably become part of the classroom. The policy was simple: teachers set the rules for their own classes, and students followed their teacher’s guidance. This flexibility allowed different disciplines to adapt AI use in ways that made the most sense for their work.
While the policy remained steady, it became clear that students needed more clarity. That’s where the AI Assessment Scale came in. Originally developed by Leon Furz, an expert in the practical and ethical implications of Generative Artificial Intelligence, the scale categorizes assignments into levels, from those where AI use is not allowed to those where AI tools may be freely used.
Academic Dean and Athey Family Master Teacher of English, Katy Hudak, who helped bring the scale to Hill, said she discovered Furz’s work through podcasts and LinkedIn. She saw its value immediately. “What we were discovering here at Hill was that students had this powerful tool but weren’t always sure how to use it. I wanted a shorthand that made sense across different classes,” she explained.
Hudak introduced the scale with support from the Academic Council, which includes department chairs, as well as Ari Baum, Associate Head of School and Carl W. Knobloch ’18 Instructor of Economics, who reviewed the system over the summer. Because the scale came prepackaged, Hill adopted it without any changes. This is the first year of use, and future adjustments are expected as students and teachers provide feedback.
For students, the biggest benefit has been uniformity. If one teacher calls an assignment “Level 2” and another teacher does the same, students know the expectations are identical. The scale also encourages open communication. Hudak further emphasized that “the most important thing is to ask questions and make sure you understand your teacher’s expectations.”
Student reactions have been mixed but thoughtful.
“At first I thought the AI scale would make things more restrictive, but it actually feels clearer,” Addie Beckey ’26 said. “If my teacher says Level 3, I know exactly what that means. It takes away the stress of wondering if I’m accidentally breaking the honor code.”
Jill Patel ’27 agreed, though she noted lingering uncertainty. “Sometimes it’s still a little confusing, especially since not all teachers explain it the same way, ” Patel said. “However, I like that the school is at least trying to keep up with AI instead of ignoring it. I think it helps us learn how to use AI responsibly before college.”
Other students appreciate the openness the scale brings. “I don’t really use AI for most of my classes, but I like that we can ask questions openly about it now,” Seoyun Hwang ’27 said. “Before, it kind of felt like a secret thing. The scale makes it feel like we’re all figuring it out together with our teachers.”
Faculty have also found the system useful in their classrooms. Arts Department Chair and Director of Vocal Music, Yoon Soo Lim, drew connections to earlier initiatives. “The color scheme immediately reminded me of a traffic light,” Lim said. “Years ago, our arts department even came up with a policy using those three colors to show what was permissible. So, seeing that again felt familiar.”
Lim added that the scale helps her guide students to deeper thinking. “There are some things you simply can’t Google,” she said “You can find an answer, but it wouldn’t be your answer. The scale helps me direct students to put their own thought and input into an activity.”
Currently, four integrated AI platforms are officially approved for schoolwork: Copilot Chat, Zoom, Adobe, and Canva. Students have already begun asking clarifying questions, such as whether ChatGPT may be substituted for Copilot, showing the value of dialogue around the policy.
The AI scale is only part of the broader conversation about technology at Hill. Device policies, classroom tech use, and even the role of social media in dorm culture all shape daily student life.
Like all policies, the AI guidelines are expected to evolve. Hudak sees flexibility as essential: “There’s a 100% chance this will change over time,” Hudak said. “The AI world is shifting so quickly that we need to adapt with it.”
One thing, however, seems certain: AI will remain part of Hill’s academic landscape. “AI is here to stay, so we have to learn how to coexist with it,” Hudak said.