There is “great potential to make us dumber,” teacher Aaron Girdner told students in his Capital High School business pathway class as he acquainted them with the new tool on a recent day. And “sometimes it outright lies to you.”
Still, he tasked his students with learning about the economic powerhouse of China through questions to “our AI friend,” the Chat for Schools artificial intelligence model purchased by Santa Fe Public Schools in September 2024.
There seems to be no realm untouched by artificial intelligence, which has crept into school districts across the world in just the last few years in hopes of reversing nationwide trends of educational decline.
This year, New Mexico went further, requiring K-2 students in traditional public schools and charter schools to take their reading assessments using an AI-powered tool, even as many districts, including Santa Fe Public Schools, have stepped back from setting firm policies on use of the technology.
The absence of clear guidance has created a Wild West of AI use in the Santa Fe district, fueling debate among teachers about its promise and its risks. Though a working group of teachers including Girdner and astronomy teacher Joshua Cantrell meets monthly to hash out those questions, a prevailing sentiment has emerged among them: Like it or not, AI is here to stay.
“I myself have serious misgivings about it. Here’s the thing: It’s here. It’s not going away,” Girdner told his students. “Our one chance for you guys to be successful in the future is you gotta figure out how to use it.”
Down the hall, in Cantrell’s class, Chat for Schools wrote a song about the cosmos and sang it to students in a synthesized female voice.
DEVELOPING GUIDANCE
Zelda Sanchez, a digital learning coach who instructs teachers on technology use at the K-8 Gonzales Community School and Amy Biehl Community School, said discussing AI tools brings “the gamut” of responses from educators.
“You have those that are like, ‘Yes, I know how. Let me just do it.’ And others that are like, ‘It’s the devil. It’s so scary,’ ” she said.
The school board has strayed away from developing official AI policy, instead charging staff like Katy Grunewald, the digital learning innovation coordinator, with developing ongoing “guidance” for educators and the curious public, informed by monthly sessions with district teachers.
At a recent meeting, the group looked at various news articles addressing concerns about environmental impacts of AI — and one report on a recent study on AI as a “cognitive crutch,” finding students who used AI on an assignment performed worse on memory tests than peers who didn’t.
One article suggested institutions create a formal policy guiding AI use, but “at this point, the [school] board has not mentioned any interest in writing official policy,” Grunewald told teachers.
Capital High teacher Scott Robbins spoke up: “That’s silly, right? ’Cause this is the same as cellphones.”
Students, he noted, will enter one classroom where cellphones are encouraged for classwork, while their next teacher will threaten to confiscate them.
While the district has a policy restricting cellphone use, Robbins said, it’s so dependent on the class, “what difference does it make?”
The monthly meetings are rife with disagreement, which Grunewald described as a healthy process in crafting policy for such a new, untested technology as AI in schools.
LIMITS AND RESTRICTIONS
In place of firm policy, AI guidance is fairly open-ended for teachers, though many tools remain restricted for students.
The popular ChatGPT, for example, is unavailable, with teachers steering students to district-contracted tools like Chat for Schools, which looks like a traditional AI chatbot but has controls for teachers and logs students’ prompts.
“Pockets of experimentation” like Cantrell’s singing AI sprout up in classrooms across the district, Grunewald said.
She and the working group plan to meet in July to assess the success or failure of certain tools to bring to district leaders before the coming school year.
The district’s contracted platform Brisk, for example, is an AI tool intended for teachers to develop their lesson plans. A few teachers praised the tool, saying it vastly shortens the time needed to create assignments by cutting down busy work.
Girdner said using a mix of AI tools alongside Brisk cuts the time of making a lesson plan from four to five hours down to just 45 minutes, giving him time to “overhaul” his lessons and consider structural changes.
But the tool has limits, Grunewald said, noting it recommends teachers upload special education students’ Individualized Education Programs to better shape its output.
Uploading the IEPs, which contain personal information strictly protected under federal law, is “a high-risk use that we do not encourage,” Grunewald said.
Data privacy remains a broader worry around AI tools. Contracts with companies like Google include provisions restricting data collection on Santa Fe students.
Grunewald said concerns also revolve around teachers using third-party tools like ChatGPT, which students are barred from using, and that imposing restrictions on teachers for certain tools would be “revisited” in the summer.
Amira, the reading tool the state adopted and mandates for K-2 assessments, has drawn criticism from teachers and lawmakers because it struggles to understand students with accents or speech differences.
Sanchez, though, said her experiences working with teachers on the tool — and her time using it as a classroom teacher — have been mostly positive.
She recounted using the tool during parent-teacher conferences to show families how much their children’s reading skills had grown between their first and last Amira test of the year.
“You go to their last Amira, and you can just hear the growth, like ‘Oh, my goodness. Look how they’ve grown … we have so much to celebrate,’ ” she said.
WHEN HUMAN IS BETTER
While teachers’ reactions to AI fall across the spectrum of support and disapproval, students in Girdner and Cantrell’s classes seemed to share a common perspective: The tool is helpful for specific questions, but there’s a limit to its powers.
“It’s helpful, especially when your work is hard,” said Damien Garcia, 15, a student in Girdner’s class. “But at the same time, I feel like you’re giving it too much power when you’re just, like, letting it speak for you … instead of you being original and using your creativity.”
Damien, who aims to become a lawyer, said he’s seen how AI has been used in the legal field but rejected the idea the profession would be replaced by AI.
“A computer just looks at the facts and just tries to do research,” he said. “But a human has feelings, and they have understanding on how laws have worked over time.”
Sara Villa, 15, also in Girdner’s class, agreed AI could be a useful tool, noting she uses it to ask hyperspecific questions she might otherwise be unable to find answers to online.
She was concerned, though, about the environmental cost.
“Sometimes when I use it, I do feel guilty,” she said, adding, “I know it’s not good for the environment.”
In Cantrell’s class, 17-year-old Angel Lopez said he also appreciated how AI could answer specific questions, but he delivered something of a rebuke to Cantrell’s songwriting exercise.
There’s a “split” in appropriate uses of artificial intelligence, Angel said. He recalled his distaste at one of his school’s earliest AI lessons two years ago: an assignment to compare poems written by AI in a Spanish class.
“Poetry is more human,” he said. “I feel like it’s better if we do it.”
© 2026 The Santa Fe New Mexican (Santa Fe, N.M.). Distributed by Tribune Content Agency, LLC.
