*names have been changed to protect identities of academically-dishonest students
Junior Henry Brooks* hasn’t written an original essay in four months.
He plagiarizes argumentative pieces in English, historical document analyses in history and short-answer responses in science. The first step to any written assignment has become a habit: typing “openai.com” into his search bar.
“Artificial intelligence makes cheating almost too easy,” Brooks said.
Hours of homework have morphed into 20 minutes of copy-and-pasting spent focusing more on Fortnite YouTube videos than crafting thesis statements. Is Brooks worried about getting caught and suspended?
He actually thinks that idea is funny: “I have AI that makes AI essays undetectable by AI-checking websites,’” he said with a chuckle. “There’s no way teachers will catch me.”
When research lab OpenAI released ChatGPT on Nov. 30 as the first viral AI chat tool, its impact on the classroom was unpredictable. Nearly four months later, the effects are clear: in an Instagram poll of 198 East students, 40% reported using AI tools to cheat at school. Meanwhile, educators struggle to determine AI’s rightful place in their curriculum and combat dishonesty.
“I can’t promise that there will be a major change by next fall,” the district’s Chief Academic Officer Darren Dennis, who is in charge of curriculum development, said. “We recognize that [AI] is not going away and have started having discussions, but nothing has been decided.”
Dozens of new AI websites, chatbots and apps are released weekly by start-up developers and companies like Microsoft and Google. Though the district has been banning new tools on school-issued MacBooks, 32% of students in an Instagram poll of 109 are aware of at least one AI program not blocked. Juniors Chase Jones* and Ayden Beverage-Calvin can list multiple.
“Right now [on March 28], there’s at least Rewriter Tools, SmallSEO Tools and Jasper AI,” Beverage-Calvin said.
Plus Jones didn’t even care when the district blocked his favorite AI tools like Genie and Quill Bot. He still manages to cheat on half of his assignments by sneaking his phone with AI apps into his lap while teachers are looking away. If that doesn’t work, he’ll wait to “complete his work” — code for “fire up ChatGPT” — until he gets home.
“After I used AI for the first time on a test, I was proud that my grade in the class went up,” Jones said. “It’s smarter than I am, gets better grades and saves time, so I’m going to keep using it.”
The only thing that stops Jones from using AI for every writing assignment is the prompts that ask about his personal life experiences that technology hasn’t gotten advanced enough to write about. At least, not yet. He’d be willing to complete 100% of his assignments with AI one day if possible, seeing almost no cons to “using his resources.”
“I only feel a little bad when I get a better grade than people who actually studied, but they should really just start working smarter instead of harder,” Jones said.
English teacher and IB coordinator Meredith Sternberg is unsurprised by students’ willingness to cut corners but doesn’t think it’s because they are simply “bad kids.” Instead, she feels the education system has partly failed them.
“To kids who are cheating with this technology: I’m sorry,” Sternberg said. “Education has clearly essay-ed you to death and made you not want to learn the humanities. So should us teachers be shocked when students find the most convenient, excellent hack ever and start using AI? No, because we made them stop caring.”
Even some more curious students — the ones morally against cheating at first — have caved to AI when under pressure. Sophomore Tara Nelson* woke up on the last day of the third quarter with a 79.0% in chemistry — just 0.5% away from a B. She knew she didn’t have time to write a multi-page extra credit essay herself with less than an hour until her morning class. So the temptation to ask AI to “write a 1.5 page essay explaining Avogadro’s number” was irresistible.
All she had to do was manually replace a few suspiciously-complex words like “paramount” with more common terms like “important,” and her teacher didn’t notice the ruse. Nelson’s grade was secured — with time to spare for breakfast.
“Before that day, I was nervous to try AI since I’d only seen it on TikTok and heard about other students using it at school,” Nelson said. “But it went so well that I’ll definitely use it again if I run out of time for an assignment.”
Science teacher Susan Hallstrom expects more students to try cheating with AI. But since most of her assignments are completed on paper during class time without access to electronics, she isn’t concerned about cheating in her class.
“I’ll probably change some assignments to make cheating difficult like by making more ‘pretend’ values for molar masses that can’t be looked up,” Hallstrom said. “If anything, I’m mostly concerned about [extra credit] reports and might have to change those.”
AI can be used for more unexpected assignments than just reports, according to Jones. Cheating tutorial videos on TikTok have taught him useful hacks like how to answer multiple choice questions in all subjects and create recipes in food class. Senior Ava Brown* and her friends use ChatGPT to complete extra credit AP Statistic assignments because it’s “just faster.” AI apps like Qanda can even do math homework for students.
Beverage-Calvin, on the other hand, uses an AI writing tool called Genie solely for inspiration and revision like fixing run-on sentences — he’s drawn a moral line at plagiarism. But he’s not surprised that tools are being used for outright cheating.
“It’s scary because I’ve played around with ChatGPT for fun, and if you ask for an ‘expert’ response, the AI gives an entirely different result than before,” Beverage-Calvin said. “Or if you ask for an ‘elementary-level’ explanation, it changes again. I don’t think the district will be able to stop people from using AI.”
Improvements to AI are getting more sophisticated monthly. ChatGPT 4, the most current version, draws from 100 quintillion sources to answer users’ questions compared to the ChatGPT 3’s 175 billion sources — over 500 times as many. Brooks and Jones have already noticed monthly improvements in their ability to cheat faster and more convincingly. To counter this, the district may update questions that students are asked from factual “yes” or “no”’s to ones about deeper ideas, according to Dennis.
“We want to make sure that students are doing their own work and thinking,” Dennis said. “I’m not overly concerned, but AI does make me nervous and we need to find a way to ensure that.”
Sternberg has already implemented changes into her IB classrooms by replacing essays with socratic seminars, book journals and customized tasks with specific directions that can’t be emulated by AI. Her students are receiving more “watch this video and look at this website then argue a personal opinion” type of assignments this year. But she can’t cut out all essays — it is English class after all.
“There’s essays that we’ve always had students write and maybe we can change some to make it harder to cheat, but [writing] won’t be going away,” Sternberg said.
Administration has yet to make specific school-wide policy changes regarding AI, according to Associate Principal Kristoffer Barikmo, though he acknowledges teachers’ concerns about dishonesty increasing.
“We don’t know what pieces to put in place yet,” Barikmo said. “Our building leadership team has been talking about [AI] and how it won’t be going away. It’s our new reality, like when calculators were introduced into math classes for the first time and curriculum was changed. The question now is, ‘How do we adapt instruction to this new tool?’”
Students who are caught using AI for writings will receive a non-negotiable zero in Sternberg’s class unless they’ve correctly cited AI as a source. The College Board has yet to publicize any changes to AP programming due to AI. But the IB program announced on Feb. 27 that citing ChatGPT in IB final projects is acceptable given that all AI-produced information is in quotes.
“When the IB program announced changes, it became clear that we can’t regulate AI away or simply tell students not to use it,” Barikmo said. “As a school, we’re looking at how to help students show that they’ve done critical thinking themselves by citing anything that was generated by a chatbot.”
In the meantime, 83% of students in an Instagram poll of 120 believe that they can get away with cheating using AI. Jones hasn’t received any zeroes, claiming: “I think I cheat good enough that no one will find out.”
Luckily for him, ChatGPT will continue correcting his grammatical errors until he’s caught — if ever. Getting away with cheating doesn’t take a criminal mastermind, he says. Just any student with basic computer skills and the right motivation.
“We can’t bury our heads in the sand and pretend that AI isn’t going to transform classrooms,” Sternberg said. “We need to figure out how to embrace AI at school in an educational way instead of as a force for evil.”
Related
Leave a Reply