The addiction begins with one simple prompt, then a click, then a solution. And then, with one submission, an unhealthy dependency is born. An AI chatbot—an algorithm trained on guesswork and probability—becomes an ever-giving oracle of easy solutions, serving as a digital crutch that replaces confidence, curiosity, and eventually a student’s critical thinking ability. The use of AI within LM has provided many students with a convenient and compliant servant that, when unchecked, is a vassal that becomes judge, jury, and academic executioner.
From English 1 to AP English Literature and Composition, AI-detection alarms have become the soundtrack for the LM English Department. English teachers have seen a great increase in the number of detected cases of students using AI, and students have seen an increase in AI paranoia-perpetuated scrutiny. In a world where AI use is becoming more prevalent, teachers have implemented AI detectors and applications like LockDown Browser. Across the English department, students are claiming false accusations after their essays have been flagged for AI. The future of such at-home, long-term writing assignments, previously viewed as essential parts of the English curriculum, has been put in jeopardy.
However, the advent of AI has provided students with great opportunities to improve their work and habits. In English classes, AI can help quiz students on book chapters and vocab quizzes as well as recommend books, allowing students to develop their writing and analysis skills. Many teachers, such as AP English Language and Composition teacher Laura Stiebitz, believe that AI can be a great study tool when used correctly.
But with these aids comes the potential for abuse. AI has quickly become a substitute for thinking when students choose the ease of obtaining quick answers over honest efforts. “I think that AI is very seductive,” said Stiebitz. “High school students are in a key developmental period where they need more experience wrestling with ambiguity and complex ideas and not getting the answer the easy way.” This shift away from independent thought to relying on AI-generated solutions decreases the ability for students to exercise critical thinking skills and creativity, directly going against LMSD’s first strategic pathway: redefining success to “incorporate creativity [and] critical thinking.” One of the most important things in English class is the ability to learn critical reading, critical thinking, and how to communicate, something that English 2 teacher Meredith Dyson believes is hindered because “AI gets in between the student and that actual learning.” A great amount of concern should be pointed toward the next generation of students who go out into the world with a hollowed sense of resilience—a quality gradually eroded as AI replaces struggles with seemingly easy solutions.
As the use of AI becomes commonplace in the classroom, a noticeable number of teachers have evolved their practices in an attempt to combat AI use, moving to incorporate restrictions on the usage of AI with the introduction of LockDown Browser, a web browser that restricts students’ ability to access other applications. Nearly every piece of writing in Stiebitz’s AP English Language and Composition classes takes place within the LockDown Browser—a purposeful shift, she says, to “limit the temptation for AI and limit the possibility for students to use AI.” LockDown Browser, however, also possesses many technological problems. “It’s kind of buggy sometimes,” says Felix Guerra ’28. “I remember [LockDown Browser] breaking multiple times during important tests.” But the mere need for LockDown Browser exposes the concerning degree to which many students now feel the need to outsource their thinking through AI. Though effective in the narrowest sense, LockDown Browser reflects a troubling status quo where learning survives only under digital lockdown.
Problems have also arisen in teachers’ attempts to detect the AI use in student essays. So far, AI detectors have been sold as protection for authenticity, but they are often too unreliable. The purpose of AI detectors is to identify whether or not something was created by an AI. However, studies from sources such as the Massachusetts Institute of Technology state that these detectors often operate on an incorrect assumption of certainty and can make false allegations easily. Students may have their work mistakenly identified as inauthentic while AI-produced work goes undetected. Many students have complained about the use of AI detectors in LM. “They disproportionately targeted conclusion paragraphs, even those that have been written in LockDown Browser,” said Danny Silverman ’27. “They even highlight running headers.” AI detectors have also led to students being falsely accused. More recently, close to twenty AP English Literature and Composition students were accused of using AI on one paper. Only after receiving zeros for their assignments did the course instructor note that many previous AP practice FRQs were flagged by AI detectors despite being written in LockDown Browser. Such an example clearly points toward AI detectors being an unreliable tool and should call into question the fairness of LM’s AI policies. LM has enacted an extremely rigid policy regarding what constitutes “AI-generated” content. Any assignment that registers twenty percent or greater on AI detectors is deemed AI, despite academic institutions such as the University of Pennsylvania citing over ten million studies concluding current AI detectors are unreliable and cannot be established as accurate. LM’s strict cutoff unfairly penalizes students and underscores the unreliability of AI-detection services to judge academic integrity. Many students continue to report feeling as though they are being dragged into needless “conferences” or needing to rewrite their work after having their original submitted work flagged by unreliable and inaccurate AI detectors. While AI detectors have varying levels of success and could be used to solicit inquiry, these systems are completely unreliable as tools for completely determining academic integrity. Furthermore, an AI detector’s judgment can be easily overturned by rephrasing or replacing AI-generated material. Using AI detectors to be the sole determinant of academic violations erodes respect between students and educators and fuels unnecessary suspicion.
The future of LM’s English classes must be based on cooperation and preventative measures, rather than control and detection. English class should not allow for the temptation of “thinking” to be outsourced; rather, students should be able to think in an environment of engagement and through participation with fellow students. Yet, AI detection and the reliance on LockDown Browser to maintain control over learning are indicative of a weak academic system that extends even beyond LM. To start, conversations about AI and how to use it should be paramount. “I recognize [AI] is a tool that is not going away any time soon, so I like to engage students in conversations about it, like what do they know, what do they think about it, so it’s less of an ‘us vs. them’ mentality and more of a community conversation, like what are you learning about, what are you discovering, and what do you think about the future?” emphasized Stiebitz. There has also been enormous emphasis on limiting students’ exposure to AI through applications like LockDown Browser, but there are also other alternatives. Instead of relying on AI detectors, using them sparingly along with having a sample of a student’s work where they had no access to AI provides much more reliable and solid evidence. Through timely, physical, and in-class projects, students can better adapt their creative and critical thinking skills while feeling free from the distrustful connotations and error-prone environment that LockDown Browser brings. In the age of ChatGPT, it is critical that the English classroom serve as a place where struggling is not a burden but rather an opportunity to grow and build on past experiences; it is critical that the classroom provide the opportunity for the development of thought as a result of the dedicated effort by students working through a process of learning, as opposed to the AI’s simple, shortcut solutions.
If LM is to fulfill their own definition of success—to transform students into capable and confident thinkers who can make sense of complexity, develop unique ideas, and partake in critical thinking—then classrooms should move to make conversation and prevent temptation. True learning takes place through active engagement, conversation between two parties, and the difficult but necessary act of thinking independently. It is clear LM must adapt, as AI is here to stay: “I think that it’s so important to keep talking about AI among colleagues and with our students,” Stiebitz said. “…[AI] is evolving so rapidly, and I really don’t want the richness of my job to be reduced to detective work.” It is clear that a multitude of LM students and teachers are concerned about AI’s luring nature and recognize it is a problem that is not going away. “I am sympathetic to kids who feel tempted to use AI, and I think if I were in high school right now, I would find it very tempting,” said Dyson. “This is bigger than any one student who uses AI; this is a game changer for what society is going to look like going forward, so we can’t really anticipate how it’s going to change our world, but it already has and will continue to, so part of me is very worried about that as a teacher, and especially as an English teacher.” LM must take responsibility for fostering a learning environment that equips students to think for themselves, rather than punishing them for the tools they encounter.
