Carly Waldman is a junior at Staples High School. She is a varsity swimmer, and has lived in Westport since she was 8 years old. She writes:
I’ve always been a good student and kid. I’ve never gone to the principal’s office; I don’t sit in the bathroom and vape; I work hard and complete my assignments on time.
As much as I sometimes want to skip my AP social studies class, I show up every day with only minor grumblings. I take pride in my work, and was incredibly happy with the essay I wrote on hate speech I had just submitted for my class. That’s why it was shocking to me when, the week of midterms, my favorite teacher called me outside the room and accused me of cheating.
It had taken me 2 weeks of research, drafting, and revising to write this piece. When I finally turned it in, I felt quite accomplished. It was one of those moments when you know you did well, where you can almost picture the teacher nodding in approval as they read it. But that moment of confidence was short-lived.
Carly Waldman
My teacher stood before the class a week later, explaining that she had been reading essays that didn’t sound like students’ typical writing. Too many felt eerily polished, structured in a way that suggested a human didn’t write them. She said she was concerned — so concerned, in fact, that she had decided to have us all submit our essays to Turnitin.com, the AI detection website our school uses – although it shouldn’t.
Turnitin.com has historically been used as an anti-plagiarism tool. The AI-detection piece has just started to be incorporated over the past few years in schools; it is imperfect and illegitimate. The site itself warns against using it to accuse students of AI-generating written work due to its fallibility. Yet, that was the sole basis of the accusation against me.
I didn’t hesitate submitting my essay to Turnitin when requested to do so, because I knew I had written the piece myself. A full month passed and I assumed everything had blown over — until 2 days before midterms when my teacher suddenly pulled me aside.
She looked upset as she told me my essay had been flagged due to Turnitin’s AI detection, that there was nothing I could say or do, and that she had already reported my name to the school as per mandated school policy.
I blinked, trying to process what she had just said. I wanted to show her my notes, drafts, and the hours of effort that had gone into my writing. But before I could say anything else, she shook her head.
I was told I would have to come in the following Monday – after 4 hours of midterm exams, and do an alternative, supervised written assignment – yet, in doing so I could only get my grade up to a 50%. If I didn’t do this, I would get a 0 on the essay I had worked on so hard.
She told me the only way I could fight this was to go through the appeals process. My stomach dropped. The appeal process. It sounded like a bureaucratic nightmare where I’d have to sit in front of a panel of teachers, defending my own work — a student on trial. The idea was infuriating. I wasn’t even allowed to explain myself before I was labeled guilty.
I left the class feeling like the ground had been pulled from under me. The stress of midterms was already suffocating, but now I had this added weight on my shoulders.
I called my parents for support, unclear about what to do and wanting to talk to them before they got notice from the school. My guidance counselor, someone I trusted, could only offer sympathy and a breakdown of the appeal process.
But what I needed wasn’t sympathy — I needed justice. I needed someone to believe me.
The next day and a half was stressful and anxiety-inducing. Could this affect my entrance to college? Would one of my favorite classes and teachers forever think I was a cheater?
After the teacher called my parents, she agreed that she needed to bring my essay to the department head and take a second look, a human look, at the piece I wrote.
My name was quickly cleared when they determined I wrote it; even a personal story from 10th grade was flagged as “likely AI-generated.”
But that didn’t erase the unnecessary stress that the situation created. The teacher suggested my parents write to administration. The head of the department and the principal agreed to a meeting with my parents and me. However, I felt like I wanted to handle this by myself so I went to the conference on my own. (That was a little nerve-wracking!)
The worst part of all of this was the fact that no one could really explain how this had happened. When I met with the principal and the head of the department, they were polite but vague. They acknowledged the mistake but didn’t offer any real solutions.
If AI detection software is this flawed, why are we putting so much faith in it? Why are students being presumed guilty based on an algorithm’s judgment?
Situations like mine shouldn’t happen again. Schools need to approach these accusations with more care. Instead of immediately flagging students as cheaters, teachers should talk with the student, ask about their writing process, and consider the evidence beyond AI-detection software. Technology is not perfect, nor is the system that blindly trusts it.
I worked hard on my essay. I knew my own words. And yet, for a time, I was treated as if my voice didn’t belong to me.
That’s not how justice should work. That’s not how education should work. If schools want to encourage integrity and prioritize mental health, they should rethink their policies on Gen-AI accusations, because I doubt I’m the first or last person who has gone or will go through this awful process.
(“06880” is proud to give Westport students a voice. They are our future. As for the future of this hyper-local blog: Please click here to make a tax-deductible contribution, supporting our work. Thanks!)
Anyone reading this spectacularly written piece would come away knowing for certain that Carly had written her own essay for that assignment. The only thing missing from the blog is the name of the
thoughtless, destructive and unimaginative teacher who put Carly through this bullshit…Artificial Intelligence can be forgiven for its error…the teacher’s lack of intelligence cannot.
I’m so glad this young woman had the courage of her convictions. It should have been that this teacher showed the same respect for this student as Carly showed for the process. Knowing that the system judging this was flawed, it should not be the one thing that determines the authenticity of her work. Hope the teacher and the administration have learned from this!
The fact that they accused you before a human spent time reviewing it makes me so upset. Carly, keep being your brilliant self. We support you! Xoxo
How about a human written apology from the teacher and the school? Such unnecessary stress placed on this student. I hope the school stops using the AI program.
Carly is a star — courageous to go public. I do feel for teachers in this day and age, caught in this inexorable force called AI, which is not about to disappear … next week, or next year
Hooray for Carly Waldman and her beautifully-composed essays. She will do very well in life indeed.
Wow this is so upsetting to hear how this occurred. Why did the teacher accuse this student and others of cheating just because maybe the students had learned from the teacher and other teachers too well on how to write a great essay. Maybe teachers should first ask the student individually what their writing process was like and if they had any notes they could show them as to how they got to the finished essay without accusing them of cheating first. Maybe they could say I found your essay extremely well written or better than past work you have submitted and I want to see how you worked on it to see how your skills have evolved. That way before jumping to a conclusion of cheating and making a student feel awful about their submissions and efforts, they can feel bolstered and then after reviewing this, if the teacher still feels as though the students cheated they can have a second human review all the information and see if they agree or not. If they also feel the students cheated then there should be a hearing or review board of what and why they feel the student cheated and the student should still be presumed innocent until proven guilty. Isn’t that how the justice system in the US is supposed to work?! Why do we as adults have to first damage a students mental well being and potentially cause extreme harm to many teens fragile mental stability before realizing they (the adult with power that a student may look up to) could be wrong! This is extremely upsetting and I applaud Carly for her strength and bravery at bringing this public. Also, congratulations on such a great essay! You should be very proud of yourself. The teacher and administrators need to find a better way to balance this issue.
I fear the long term impact of using such software. Will the best students be tempted to make factual or grammatical errors to avoid the false accusations this student was forced to address?
I applaud Carly for standing firm and addressing the false accusations head on. I do wonder how many other students also may have been falsely accused by the school’s use of this software but did not have the courage to seek redress as did Carly.
Kafka-esque
What a nightmare for this girl. Like life isn’t hard enough without this big brother is watching nonsense.
What incredible courage it took to write and share this – thank you, Carly! The school’s administration has a long way to go with its vaunted AI pilot. Sounds like they first need to disarm the teachers. I don’t want to jump to blaming the teacher, however – this is how they’ve been told to handle this new technology, by using (ironically) an AI tool (Turnitin) uncritically to punish students.
I asked ChatGPT to write a short piece about a High School student being accused of using AI to write an essay that she alone wrote. Here is its answer.
The Essay Controversy
Sophia Reyes had always been a diligent student, known for her insightful essays and thoughtful analysis in English class. So when she turned in her latest paper on Frankenstein, she expected her usual feedback—maybe a note about her strong thesis or a suggestion to refine her conclusion. Instead, she was called into her teacher’s office and accused of using AI to write it.
“This doesn’t sound like you,” Mr. Donovan had said, holding the printed pages like they were evidence in a trial. “The vocabulary, the sentence structure—it’s too polished. Did you use an AI tool to generate this?”
Sophia’s stomach dropped. She had spent hours on the essay, meticulously crafting her arguments. “I didn’t,” she insisted, heat rising to her face. “I wrote it myself.”
But proving her innocence was harder than she expected. The school had recently implemented AI-detection software, and her essay had flagged as “highly likely” to be AI-generated. She protested, argued that she had simply improved her writing style, that maybe the detector wasn’t perfect. Yet, the suspicion lingered.
The irony wasn’t lost on her—she had written about Frankenstein’s monster, a creation judged before being understood. Now, she was fighting against an accusation she couldn’t disprove, forced to prove her own humanity in a system that no longer trusted it.
Speaking generally, many of us use editing tools, like Grammarly, which now use generative AI enabled by default for edit suggestions. If the edits are followed, the writing could be flagged.
Putting students with AI-detected writing through the plagiarism pipeline is way too severe a punishment for “evidence of cheating” which cannot really be substantiated. A conversation is at least warranted first, especially if this is unusual for a student.
If Westport Schools administration forcibly outsources all their teachers judgement to technology, eventually the former won’t really be needed as part of the “education” process.
It occurs to me that one obvious outcome of all this will be an AI model that takes a draft paper, human or machine-generated, and iteratively runs it through turnitin making subtle changes such as tiny otherwise unnoticeable flaws until it “passes” scrutiny. The model will be trained by feeding it masses of A+ human-generated papers as well as the collected ones failed by turnitin.
Or, like in the good ‘ol days, just have your parent write it for you.
Small grammar error above intentional. I am not a robot.
My stomach was sick imagining her in moment outside the classroom.
I know how my dayghter would have felt. I also can’t imagine how distracting the stress of this was during the exam week. It could have all waited until after that. I hope the school, and I have shared with with my HS, put policies in place that seem more fair and less accusatory at the get go. I feel so bad, but also see the grit this kid had and grew standing up for herself and others. Thank you for sharing your story!
As a professor, I can say that the amount of cheating with AI that I see is disgusting. You often can tell when students cheat. It is sad when you assign a problem and you have 10 students turn in a slightly varied answer to what ChatGPT says. I have tried changing my prompt to ask for the answer to relate to personal experiences and have literally gotten answers back that say, “As ChatGPT I have never experienced that, but if I had…”
All that said, I will discuss with the student and ask them questions. If they know nothing, I know they cheated. If they claim innocence, I tell them I need to see their AI browser history for any and all browsers I choose.
It is important to speak with the students and get to the truth. I did have a student show me that he wrote an essay and then put it into ChatGPT to “make it sound better.” The result was that he had written it, but it did not sound like his work. I learned that for future classes I told students my class was not an English class and I wanted them to do their best, but I was most concerned about the content related to the material I taught.
Saying these tools should not be used is just not practical. I caught two students who turned in identical papers they had evidently purchased from the internet. There have to be consequences to cheating and better the students learn them when they are young and it won’t impact them as much as letting it go and having a major issue later.
Dear Carly-
Your 06880 letter should be the introduction to the Turnitin website as required reading before submissions.
AI is like an untrained dog that can bite you unexpectedly and it’s owner thinks you did something to provoke it. That said, teachers that are well-meaning, thoughtful, and diligent to flag your paper often aren’t sure what to think when they have an exceptional student that does work well above their grade level or are very creative (been there). In this case they chose the wrong not ready for prime time ‘tool’ to figure it out. As they say “what doesn’t kill you makes you stronger”, your letter and approach is proof of that, bravo young lady!!!
Turnitin.com has been in use for more than a decade… students and teachers have been feeding Turnitin’s large language models with original copyrightable papers for years. What is ironic for Carly’s situation is that her essay is now part of that large language model… AI venture capitalists are making their profits off our children’s work, and no doubt the school system is paying them a fee to do it.
This specific AI hallucination is typical of what we will increasingly see as big technology is deregulated and put in charge of more of our work and activities.
Carly’s mom here to say that this isn’t a public complaint, as one snarky reader suggested. Her intention was to bring something to light that hasn’t previously been discussed in any public forum in westport in hopes of preventing this from happening to another innocent student. This is not one teacher’s policy, so Carly doesn’t place the blame on her; this is a school-wide mandate. The principal and head of the department talked to Carly about forming a committee, that would include students, to take another look at this policy and process but it’s been a month…silence.
Um… because she didn’t do anything wrong, so why should she just take the punishment instead of defending herself? Also, this is a story that should be told by a wrongly accused, clearly strong, young woman? You seem to be suggesting that when someone is treated unfairly they should just roll over and take it… why bother trying to correct/improve the situation. What a sad world this would be if nobody stood up for themselves or for others.
I fully support the school creating a student committee on AI. I would bet students know more about how AI works and is changing than most adults in this town. What is great about AI is that it is learning. What will challenge educators the most is that AI is learning. If any students put two or three of their papers / ones they wrote themselves / into AI, AI will learn their writing style. Accusing students of cheating using AI will be a woeful waste of time. It is time to get creative and stay ahead of the technology and help the students harness it. students might suggest to teachers to flip the assignment. Use AI at home to research a topic. Use the old fashion blue books in school to have students handwrite an essay on the topic.
Carly is our next-door neighbor. We’ve watched her grow up. Unnecessary though this incident may be, Carly will learn from it and, as she had done for all the years we’ve known her, it will make her stronger. As a teacher, I know how difficult it is to judge a student’s work and now, with the advent of new technology, I can’t imagine how much more difficult it is to effectively teach this critical skill. I see this incident as a watershed moment for all stakeholders in educating the next generation of writers. I’m proud to say I know one of them and, from how she rose to meet this challenge, success is emblazoned on her future.
I echo Joan Gillman’s comments. AI is important to the future of everything, so learning to use it wisely and responsibly will be important for students and teachers/administrators alike. But it should not be implemented in a way that allows either students or teachers to abdicate responsibility for learning, teaching and evaluating. AI does not inherently have any sense of equity, fairness or human judgment (at least not yet). The administration should not rely on AI to evaluate students any more than students should rely on AI to submit work. Kudos to Carly for the courage to speak out publicly on the topic.
It is incredibly disturbing to read Carly’s account of what happened. I don’t know her, but I am so glad she brought this to the attention of the community and think she is very brave for doing so. Does anyone at the school or the district have a true understanding of how these “academic dishonesty” tools work? I highly doubt it. Detecting plagiarism is one thing, but training a model to detect if something was written by a human versus a machine is quite another. I can only imagine how many false positives there are which upend a student’s life. The district should re-think its policy of utilizing a tool that is inherently inaccurate to label a student’s work in this way.
The school faculty and administration have obviously been skipping their homework for quite some time. Less public posturing on policy as a means of obfuscating professional negligence and more listening to the students would be a good remedial program for them to pursue. How anyone could question this young lady’s integrity is beyond comprehension. It is very regrettable that it was necessary to air the matter via social media which seems to be the venue of choice for all things Westport.
How awful and kudos to the student! Teachers ( and college professors) are put in a bind these days with the rampant use of AI, but there’s really no excuse for not assuming the student was telling the truth and taking reasonable steps to verify this, rather than putting her through this stressful nightmare as if she were obviously the one at fault. I’m astonished that the teacher and school didn’t seem to have a more reasonable and personal protocol to address this. I’d like to hear from the school as to what can be done in the future to avoid relying foremost on software instead of common sense in reviewing students work.
Thank Heaven we have students like Carly with the
Fortitude to persevere in the face of lazy injustice.
Brian Strong
There’s something bizarre about using AI intelligence to determine if a child’s writing was created by…AI intelligence! Come on people, if you want students to write without tech then lead by example.
Kudos to the student for publishing and great job on the narrative. Too many hear “outraged” over what happened but the reality is AI already can be used to generate essays and other homework assignments, and you’re kidding yourselves if you believes students aren’t taking advantage of it. It’s only going to get worse as the AI models become more effective and schools & universities will have their hands full. Sadly, ones who lean on AI to do their work will only cheat themselves and find out the hard way as young adults. The lesson here is schools & teachers will have to be quicker and more thoughtful in how they “police” AI in assignments, and parents should perhaps engaged even more with HOW their kids are getting their work done.
Carly: You did something incredibly brave and community-minded by publishing this piece — you could have put this unfortunate incident to rest once your own individual situation was “resolved” but instead you are doing the hard work of making sure processes and systems change for everyone at SHS. Brava!
And parents, this is the same district that is about to implement an AI “pilot program” using sixth graders and Staples science students as the test subjects. This gives me… pause.
Thank you, Dan, for printing Carly Waldman’s letter. You’ve given this intelligent, courageous young woman a voice that the community can hear.
Anti-Semitic Incident: An Alternative Perspective
Having done a casual survey of my peers, I think I speak for a significant portion of Westporters in thinking that the correct adult response to finding an offensively designed snowman in the forest, would have been to remove the mustache and emblem, and then continue my walk while silently lamenting the state of today’s youth… It would not have been to take pictures and then publicize as widely as possible something that was clearly a teenage prank.
Call me naive, but I am 99% confident there is no secret cabal of neo-Nazi activists hidden amongst us hatching nefarious plots, against whom we must launch a campaign of vigorous discovery. Somewhere in Westport are a couple teenage boys high-fiving in the basement at their success in scoring a trolling tri-fecta, eliciting public responses from the town administration, the media, Team Westport, the ADL, the synagogues, et.al.
Having once been one, I can attest that teenage boys are a demographic not distinguished for their thoughtful consideration of others. Like little boys testing out ‘bad’ words on the grown-ups, the teenagers know the hot buttons of the day which they can push, simply for amusement, to get a rise out of the authorities. Over a beer, I’d be happy to share some questionable teenage behavior of my own.
I suggest the right response when ‘a tree falls in the forest when no one is there’ is not to amplify the barely audible sound. Let it be. Don’t give oxygen to pranks, or trolling, by launching ‘anti hate’ campaigns against ‘haters’ who are, I suggest, effectively non-existent in Westport. This kind of response fits perfectly into the MAGA gameplan, which is to trigger the ‘bien pensant’ of the left to mount their soap boxes to harangue the unwashed with lectures on what heathens they are. I for one am so horrified and disgusted by the democratic loss to Trump that I want to win in 2028, not just feel good about myself by preaching to the choir.”
From God’s mind to yours and thus to ours.
FYI Turnitin has been in use in graduate, undergraduate, secondary, muffle and elementary education for more than 25 years.
If any AI application has been thoroughly vetted, it is Turnitin…
AND YET it still makes serious mistakes that in another situation could seriously affect a young person’s record and future prospects, impair their reputation, etc.
MORE it feeds on what we create.
Today JD Vance demanded in a world wide conference on the future of AI that it be totally deregulated. Since he is a billionaire venture capitalist, I completely understand his point of view.
I opt to stand with Carly today and the future victims of vulture capitalism and AI in the future.
This “educator” is lazy, stupid, and lacks empathy and trust for this young woman.
Suggest some form of administrative action be taken against such a disappointing individual.
Robert, just to clarify, this was not the teacher’s decision, this is a school-wide mandate and process that the teacher had no control over. The teacher is wonderful and remains one of my daughter’s favorite teachers as well as one of her favorite classes. She was just doing her job, it’s up to the school to change the policy.
I’m glad I don’t have kids in school these days, if this is representative of what families have to deal with. I work in AI, and I am appalled that the school/district allows this. I encourage parents to protest this use of technology in this application – at least, with the apparent lack of governance they seem to have adopted. The comment above about turnitin’s long history is an example of technical illiteracy that threatens us from all sides. The technology used to classify text as “likely AI-generated” is *completely* unrelated to any technology (e.g. text is “likely plagiarized”) deployed before November 2022. The idea that because a company has been working on technology X for a long time, so we should trust the unrelated technology Y they just recently deployed is not even wrong. Bear in mind the irony that the new LLM-based approaches are built on stealing the writing of authors and rights owners all over the world. It’s pretty rich for that toolset to then be used to accuse someone of plagiarism without any recourse.
appreciate m the clarification.
So who is ultimately responsible for this policy implementation?