Artificial Intelligence (AI) in the classroom has been a big conversation this back-to-school season. How are districts across the country dealing with the increasingly popular and accessible tools?
The two largest districts, New York City and Los Angeles, have banned AI in most classrooms.
On the other end of the spectrum, the Alpha School, a K-12 private school that replaces traditional teachers with AI, is rapidly opening campuses across the country. It’s trying to operate in Lexington, but needs the local School Committee’s approval. Alpha likely won’t get a vote from the committee until December, Lexington Public Schools superintendent Julie Hackett told the Observer.
So what are Lexington’s guidelines for teachers and students at LPS?
It takes some digging, but there are two documents on the LPS website that seem to be the district’s stated policy: “Guidelines for the Use of AI in LPS” and “AI Use in Lexington.” But after checking in with Thomas Martellone, director of innovation and instructional technology at LPS, LexObserver learned that those documents don’t exactly stand.
About three years ago, LPS digital learning coaches (seven staff members spread across the schools who would visit classes to teach teachers and students how to use technology responsibly) went to an AI conference. Based on their findings, the district made a first pass at guidelines. What’s on the website now is a second iteration of that first attempt.
But now, the district is updating them again to keep up with AI’s evolving capabilities.
Over the summer, LPS administration had a meeting with Russ Wilcox, an expert on AI, where he helped staff “think through how [they] can partner with AI in schools.”
“The professional developer was talking about how [teachers could have students] use blue books and then [teachers] could have AI help then read the responses from the blue books and create critical feedback for kids that human beings can’t generate so quickly, which is an interesting thought,” Hackett recalled. She detailed more takeaways in a memo she sent to LPS families on Aug. 25.
Hackett’s general feeling is that the school district should “lean into the benefits of [AI] while understanding and appreciating the dangers, help kids think about how to use it to enhance their critical thinking skills, and learn how staff can use it for efficiencies so they can further their work in more advanced ways.”
“People have strong feelings about the environment, climate, and then there is worry about what personal data and information are going to get out there from the use of it, but we also have to be responsible and help students understand how to coexist with it in this changing world,” she added. “I don’t think shutting down anything related to AI is the answer. I think it’s really complicated and we have to be thoughtful about it.”
In the meantime, teachers are allowed to use AI to plan lessons and for efficiency purposes. Martellone vetted which tools are safe to use in terms of data privacy — the district has data privacy contracts with Canva and Grammarly. Staff also have access to Gemini and other forms of assistive and generative AI, but aren’t allowed to input any confidential or private information about students into AI.
Students can’t access any AI tools on their school-issued devices (at least, theoretically — we all know kids find ways).
While LPS administration considers new guidelines, some teachers feel like they’re in the wild west.
“I wish we had a more coherent and consistent policy across the district,” Howard Wolke, a sixth grade ancient civilizations teacher at Diamond Middle School told the Observer. “We really need to make sure people are aware of the downsides of AI. There are obviously the environmental downsides, but also…there needs to be more transparency about it, more understanding of why it may not always be the best option.”
Wolke does not use AI as a teacher and does not allow his students to use AI, not that they would be allowed anyway — just about all sixth graders are under 13 and most major AI platforms, like ChatGPT, Claude, and Gemini, are 13-plus. Wolke teaches his students that “AI is not able to be creative — it can create, but not be creative, and they can all be creative.”
Wolke worries there is a disconnect between LPS teachers’ opinion of AI-use in schools and the administration’s.
“When you talk to most teachers, as far as I’ve been able to tell, at the middle and high school level, they tend to not be this gung-ho about AI for students,” he noted.
Wolke has been teaching at LPS for about 19 years. For his first six, he taught social studies. Then, he became a Digital Learning Coach (DLC). In that role, his job was to help educators and students learn how to use technology, including how to use AI ethically.
Due to budget constraints, and to Wolke’s disappointment, the district cut the DLC program in the spring to help balance the fiscal year 2027 budget. Now, about 13 years after becoming a DLC, Wolke is back to social studies, teaching ancient civilizations.
But Diamond teachers have leaned on his AI expertise this year. He created guidelines that his students all signed and must follow. Another team of students at Diamond adopted it, too.
“I helped two of the sixth grade teams, last year, craft policies. We made an AI contract for the kids to reference, I have taught my students about it, and we’ll keep reviewing it over the course of the year,” he said.
“Students feel like they need to use AI because they don’t have time to complete an assignment or they don’t understand it. We, the teachers, will always work with them and help them, give them more time if necessary — that’s way better than using AI for that purpose.”
Wolke does not know what other teams or teachers across the district are doing, though.
One of the only teachings within LPS that addresses AI head-on is Melissa Soule’s English elective, “Media of the Moment,” that LHS seniors can take. But because it is an elective, not all do.
In “Media of the Moment,” or “MeMo,” as it’s nicknamed, Soule talks about how to use AI responsibly to prepare seniors for the real world where AI is everywhere.
“Sometimes we are intentionally using AI to figure out when it is and isn’t reliable or helpful for certain types of work that they might be engaged with,” Soule told the Observer.
Soule generally does not find generative AI to be appropriate for all but high school seniors.
“I think it’s an age and autonomy question,” she said. “When we’re talking about generative tools, it’s generally not age appropriate because a student shouldn’t be able to generate what they can’t do themselves. They have to have those foundational skills first.”
“I’m continuously engaging in a conversation around, ‘next year, when you’re at college, you have to make these choices for yourself, you’ll have to decide where you morally land on this,’” she said.
In addition to Soule’s class, Martellone and LHS principal Andrew Baker are working with LHS librarians to create a two-part AI lesson that librarians will teach to all ninth-graders. Their goal is to have that delivered to freshmen sometime this year.
“The librarians are probably, in terms of our remaining staff — they know about media literacy, research, things like that. They’re fairly well-placed to pick up some of that material and make sure that students are still operating some baseline education,” Martellone said.
In addition to MeMo, Soule teaches sophomore English. She does not use AI at all in that class.
“It’s really key that students demonstrate core skills before they start automating them because they need to know how to evaluate the content that AI is spitting back out and they can’t do that unless they can generate the content themselves,” she said. Soule discusses the importance of building foundational critical thinking and analytical skills with her sophomores so they can potentially use AI tools in the future.
Soule said LHS’s English department has some AI guideline language in its course expectations documents. She does not know whether other departments distribute the same type of language to students.
Because AI is so easy to access, Soule is moving away from using online material in the classroom.
“I do a lot of ensuring my materials are differentiated, delivering them to students in a non-digital space, we are really engaging with each other in the classroom, a lot of stuff is printed out, a lot of stuff is hand-written, I’m trying to do as much face-to-face feedback as possible.”
Soule added that she isn’t sending home much homework beyond reading because “any open-ended question, any writing task, any non-hand-drawn creative work, we were just seeing all that get generated because the kids are strapped for time and they’re stressed out and it’s just really hard to say ‘no’ in that situation…it’s a huge temptation.”
Rory Burton, a senior at LHS, and Jiawei Liu, a sophomore, think students should be taught to use AI in moderation to curb that temptation.
“You don’t want to completely ban something so when kids get ahold of it, they rely on it too much or they overuse it,” Burton said. “Students should be taught how to use it but they should not be allowed to use it for course work.”
“Students, especially in the young adult phase, when people tell them not to use something, they might actually use it more,” Liu said in a separate conversation. “Students should practice not going to either extreme of either not using AI at all or using it too much. The best way for schools would probably be to teach students what the best usage of AI would be…it’s not going to be possible for students to just not use it at all.”
Asked if there is a big issue of students cheating with AI, LHS principal Andrew Baker said yes, but issues of academic honesty aren’t new.
“Almost all high schools have encountered [cheating with AI] on some level,” he said. “There’s not really a fool-proof technological way of knowing whether a student has used AI.”
A teacher might see a student using an AI tool on a personal device in the classroom even though students aren’t allowed to use personal devices during class, Baker noted. Or, if a teacher suspects a student used AI, they might follow up with them and ask them to talk through their thought process. If the student used AI, it would then become pretty apparent when it’s difficult for them to describe how they arrived at a thought, the principal said.
But Baker thinks teachers are more concerned about cognitive offloading than cheating.
“I think teachers really start to worry about if our students are using this to substitute for the higher-order thinking or skill that we’re asking students to develop.”
Neither Liu nor Burton said they use AI to complete schoolwork. But in his freetime, Liu uses AI to write code to create gaming apps. He noted that his teachers avoid it, too, but he thinks it “should be used to cut all of the mundane things that teachers have to do.”
“I think it’s good for organizing materials and grading tests,” he said. “Most professions nowadays can use it to help cut out repetitive work, but I think there is a clear line where AI starts replacing a teacher’s own judgement.”
Burton, who likes to write in her free time, is more of a skeptic.
“I hate what it has done to creative spaces and what it has done to the planet so I try to avoid it at all costs,” she said. “it’s just not something I feel is an ethical tool.”
Wolke found that his middle schoolers have a similar perspective.
“One beacon of hope I have is that many students recognize the problems of AI,” he said. “This generation doesn’t like seeing AI images or videos, it turns them off more often than it turns off older generations, which, to me, is a good thing.”
Burton thinks teachers should stay away from using the tool because it produces a lot of inaccuracies and interrupts relationship building.
“I feel like at that point (that teachers use AI to teach), I could go home and be taught by my computer,” she argued. “I feel like at school, a lot of it is human connection and what specific teachers have to offer.”
While Soule is against younger students using these tools, she’s aware that a lot of the conversations being had about AI can feel very catastrophic.
“I think that’s something that weighs very heavily, especially on students that are so bright like so many Lexington students,” she said. “They pay attention, they worry about the world, they want to make it a better place, they’re wondering, ‘am I going to be able to exist in the field I want to be in by the time I get my degree?’”
So she finishes the year on a techno-optimist note by pivoting students away from the doom and showing them what else the future could hold.
“I try to take us toward a, ‘who’s talking about a future that sounds better than this? Let’s look at some other philosophies of how the future could go, let’s look at some other options of what we can work toward collectively and individually. I think they need that,” she said. “I feel that we have a responsibility to young people to help them envision the world being a better place.”
As a next step in creating a new AI policy for the district, Hackett hopes to host a town hall where she and other administrators can hear from the community. She’ll send more information on that to LPS families sometime in October.
“We just have to think about how we can leverage it responsibly and keep in mind all of the things we care about: social justice issues, big data centers, the impact on the environment is such a deterrent, and at the same time, there are benefits,” said Hackett. “I want to engage the community in a conversation about all that and then figure out how we can create a governance structure for Lexington.”