When a Hingham High School teacher took a close look at the script two of her students had turned in for a history project, she found a citation to a book that did not exist. The “source” had been conjured out of nothing — one of the telltale fabrications generative AI produces when it’s asked to sound authoritative. The students, it turned out, had copied AI-generated text into their work without checking any of it.
What followed has become one of the first real legal tests of how schools should handle artificial intelligence. According to the Associated Press, Dale and Jennifer Harris sued Hingham officials in federal court after their son was disciplined over that project — a short documentary about the basketball legend and civil rights figure Kareem Abdul-Jabbar. The parents argued the rules around AI had never been made clear. The school argued he had cheated and failed to disclose it. In November 2024, a federal judge sided with the school, at least for now.
The dispute is bigger than one student, one teacher, or one assignment. It’s a preview of an argument now playing out in kitchens and classrooms everywhere, as AI moves into homework, research, writing, and test prep far faster than school rules can keep up.
What happened in Hingham
The project was part of National History Day, a long-running academic contest, and the two students — juniors at the time, in an AP U.S. History course — set out to make a short documentary about Abdul-Jabbar. By the account in the court record, AI wasn’t off-limits entirely. Students were allowed to use it to brainstorm topics and track down possible sources. What they weren’t allowed to do was have it write the thing for them.
That, the school concluded, is close to what happened. When the teacher, Susan Petrie, ran spot checks on the draft using Turnitin and a revision-history tool, the writing flagged as AI-generated, and the endnotes pointed to books that didn’t exist — including one credited to an author listed only as “Jane Doe.” The students had, in the words of the eventual court ruling, “indiscriminately copied and pasted” text from Grammarly’s AI tool, hallucinated citations and all.
The consequences were real but limited. The two students received failing grades on parts of the project and were allowed to start over separately; the Harrises’ son earned a D on his second attempt and a C-plus in the course overall. He was given a Saturday detention, and that spring he was passed over for the National Honor Society. (He was later allowed to reapply, and was admitted.)
His parents sued. Their core argument was that the student handbook said nothing specific about AI at the time, so it was unfair to punish him for it. They asked the court to raise his grade to a B, wipe the discipline from his record, and bar the school from calling what he did cheating — noting that selective colleges he hoped to attend, including Stanford, don’t consider applicants with disciplinary histories. The school saw it differently. In a court filing, officials called the discipline “relatively lenient and measured” for the unauthorized use of AI and the failure to cite it, and said the student had cheated and plagiarized. The chair of the Hingham School Committee declined to comment publicly, citing student privacy and the active case.
What the court said
The legal fight didn’t end there. In November 2024, U.S. Magistrate Judge Paul Levenson declined to give the family the emergency relief it wanted, refusing to order the school to raise the grade or erase the detention while the broader lawsuit proceeded.
Levenson acknowledged that generative AI poses genuinely “nuanced challenges” for educators. But he found that the school’s existing rules against passing off someone else’s work as your own were enough to put students on notice, and that the evidence didn’t support the son’s claim that he had simply been confused. The student had testified that the rules felt unclear because AI was permitted for generating ideas and finding sources. The judge wasn’t persuaded. The record, he wrote, showed the students hadn’t merely used AI to explore topics — they had lifted its text wholesale and never checked the sources it handed them. In the student’s own words, recorded months before the lawsuit, the expectation had actually been clear to him: if you use AI, say how and why, and be honest about it.
For parents, that distinction is the heart of the matter. The court did not rule that using AI is cheating. What mattered was how it was used — whether it was authorized for that task, whether the student copied material instead of producing his own, and whether he disclosed it. The case remains open beyond this early ruling, but on the central question, the judge found the school had acted within its rights.
Why this matters beyond one school
If Hingham’s rules turned out to be clearer than the family argued, the wider picture is genuinely murky — which is part of why the case struck such a nerve.
Student AI use has climbed fast. The Pew Research Center found that 26 percent of U.S. teens ages 13 to 17 reported using ChatGPT for schoolwork in 2024, double the 13 percent who said so a year earlier. And teenagers themselves draw a sharp line about what’s fair: 54 percent told Pew it’s acceptable to use the tool to research a new topic, but only 18 percent thought it acceptable to use it to write an essay.
That gap is where the arguments begin. A student figures research help is fine; a teacher sees the line crossed into writing or copying; a parent assumes the rules were vague; the school assumes its honesty policy already covers it.
A 2024 Common Sense Media report, “The Dawn of the AI Era,” captured how much of that confusion is baked into the system. Forty percent of teens said they had used generative AI for schoolwork, and 46 percent of them had done so without a teacher’s permission. About six in 10 said their school either had no rules for generative AI or that they weren’t sure whether any existed. And 83 percent of parents said their school had never communicated with families about it at all.
That’s the real problem the Hingham case points to. The tools are already in students’ hands; what’s missing, in a lot of districts, is a shared understanding of what’s allowed. Tellingly, Hingham itself rewrote its student handbook to spell out AI rules after the dispute — exactly the kind of fix that’s easier to make before a conflict than during one.
AI can help — but it can also hide learning
None of this means AI is just a cheating machine. Used with some discipline, it can help a student brainstorm questions, work through a confusing concept, sketch out a study plan, practice vocabulary, check grammar, or pressure-test a shaky argument.
The risk is subtler than a robot writing an essay. It’s that a student skips the very work that builds understanding — the reading, the questioning, the organizing, the drafting and revising — and ends up with a polished page they can’t actually explain or defend. A teenager who asks AI to walk them through the causes of a war may be learning a great deal. One who pastes its paragraphs into a final draft, unread and unchecked, is doing something else entirely, and, as the Hingham students discovered, may be handing in invented sources without even realizing it.
The line worth teaching a child runs right there. And it’s worth drawing before an assignment is due, not after.
What clearer rules actually look like
UNESCO, in its guidance on generative AI in education, makes an obvious but underappreciated point: these tools are evolving faster than the policies meant to govern them, and schools need clear, human-centered rules for ethical use. In practice, that means a blanket warning against “cheating” often isn’t enough. Students need to know where the line sits assignment by assignment.
The clearest teachers say exactly that, up front. For example:
- No AI on this one.
- AI is fine for brainstorming only.
- You can use AI to find research questions, but the writing has to be yours.
- AI grammar feedback is allowed; the final wording must be your own.
- If you use AI, cite it — and verify any source it gives you.
- Hand in your notes or drafts so your process is visible.
The flip side matters too, and students should hear it just as plainly: “the teacher never mentioned AI” is not a loophole. Copying work, inventing sources, hiding outside help, or turning in words that aren’t yours can violate an academic-honesty policy whether or not the handbook names the specific tool — which is essentially what the Hingham ruling found.
What parents can ask before the next assignment
Parents don’t need to master the technology. They need to ask a few sharper questions before a big assignment lands. A short version, to put to your child:
- Did the teacher say whether AI is allowed — and for which parts: brainstorming, research, outlining, grammar, or writing?
- Do you need to cite it if you use it?
- Can you show me your notes and drafts?
- Did you actually check any source the AI gave you?
- Can you explain the final answer in your own words?
“Everyone uses it” is not an answer; ask what this teacher actually permitted. And “the teacher didn’t say” is not a green light; it’s a reason to ask before turning anything in. That single habit — asking first — heads off most of the trouble.
What students should know
For students, it comes down to a simple principle: AI can assist you, but it can’t replace you. A few rules follow from that. Don’t copy and paste AI text into an assignment unless a teacher clearly allows it. Don’t trust an AI source without confirming it’s real. Don’t hide AI use when disclosure is required. Don’t hand in anything you couldn’t explain on your own. And if you do use AI, keep a record of how you used it.
The aim isn’t to scare students off a powerful tool, but to help them use it honestly — and to spare them from learning the hard way, as one Hingham junior did, that AI will happily invent a source and let you turn it in.
The real lesson for families and schools
The Harris case is, in the end, a story about a gap — between how fast AI arrived and how slowly the rules caught up. Closing that gap takes more than punishment. Schools need policies specific enough to act on. Teachers need support and concrete examples. Parents need to hear from the school before a problem starts, not during one. And kids need to absorb the idea that AI is a tool to think with, not a way around thinking.
The line between using AI to understand your own work and using it to dodge the work entirely won’t always be easy to see. But where schools don’t draw it clearly, students and families will keep finding it the hard way, one disciplinary meeting at a time. For 3A1Z Stories, that’s reason enough to put this conversation at the center of the parent–school relationship now, rather than waiting for confusion to harden into discipline.
So before the next project begins, it’s worth asking the teacher, the school, and your own child the same plain question: What are the AI rules for this assignment? It’s a small thing to ask. It can save a family a much harder conversation later.
Quick parent checklist
- Ask whether your school has a written AI policy — and whether teachers set rules per assignment.
- Don’t let “the handbook didn’t mention it” stand in for “it’s allowed.”
- Teach that AI can help with ideas but shouldn’t replace original thinking.
- Warn your child that AI routinely invents fake sources and citations.
- Tell them never to use AI secretly when disclosure is required.
- Have them keep drafts, notes, or screenshots of any permitted AI use.
- Make sure they know how to cite AI when it’s required.
- Push them to ask the teacher before using AI, not after.
- Remember that “research help” and “writing help” are not the same permission.
- Ask how the school makes — and lets families appeal — AI discipline decisions.
