The Sequoia Union High School District officially adopted MagicSchool for the 2025-26 school year after a pilot program, as Bay Area schools deepen their use of artificial intelligence tools in classrooms. The decision places one of the country's dominant tech hubs at the center of a national debate over whether AI helps students learn or erodes critical thinking.
Barbara Reklis, Director of Instructional Technology for SUHSD, confirmed the district piloted MagicSchool and Brisk Teaching before selecting MagicSchool for full adoption. The platform offers lesson plan development, rubric creation, content generation, reading personalization, and administrative reporting. A companion student interface breaks complex assignments into smaller steps and delivers AI literacy lessons.
John Rowe, a business and marketing teacher at Carlmont High School, used the platform to let students quiz themselves on uploaded curriculum. "Not all the questions were great. And then it would hallucinate," Rowe said. Students reported that MagicSchool sometimes marked correct multiple-choice answers as wrong, while accepting incorrect fill-in-the-blank responses. Still, Rowe remains optimistic. Data from the Digital Education Council supports that view: 86% of higher education faculty in 2025 said they would use AI for teaching in the future.
The push to make studying easier than cheating
Ryan Trattner, co-founder and chief technology officer of AI for Teachers platform StudyFetch, said the company interviewed over 100 students while building its software. The core function lets students upload materials that an AI tutor called Sparky teaches back to them. The platform also generates study aids and captures live lecture notes.
"Everybody is using AI as a writing or cheating tool, but we wanted to do something that actually helps education," Trattner said. "We're huge believers in this overall concept of StudyFetch making learning easier than cheating. You have to think critically in order to learn. We wanted to make studying easier for people, easier to get started, and more engaging without degrading the process of what the student learns."
SUHSD is currently revising technology policies for staff and students to accommodate AI, Reklis said. Trattner acknowledged the company has held many conversations with schools about what types of software are acceptable supplements to classroom instruction.
Teachers split on AI's classroom role
Not all educators support the shift. According to Whiteboard Advisors, 35% of high school teachers view AI tools negatively, and 25% of teachers overall say they do more harm than good in K-12 education. Connor Fenech, who teaches AP U.S. Government and Microeconomics at Carlmont, sees a direct conflict between AI use and the purpose of his courses.
"The classes I teach are about thinking, problem solving, and expressing an idea," Fenech said. "To rely on AI to do that, there's a lack of humanity, and it takes away from you being able to claim ownership. Learning is to meet something head-on and go, 'That was difficult,' or 'That got me thinking.' AI takes that away."
Nearly 60% of surveyed educators told NPR Illinois that AI is eroding interpersonal trust between students and teachers. Trattner agreed the human connection in schooling is critical, pointing to the failure of remote learning during COVID. But he believes AI can replace certain types of instruction, particularly large lecture-hall college courses where students do not seek one-on-one relationships with professors.
Rowe went further. "It's accelerating at such a fast pace that teachers will be needed less for the teaching aspect," he said. "Maybe teachers are more like a supervisor to make sure there's someone in case of an emergency. The teaching could potentially all be done inside the technology itself."
Why this matters for educators
District-level adoption of AI platforms like MagicSchool signals that tools once tested in isolation are becoming institutionalized. For teachers, that means technology policies, grading workflows, and lesson planning expectations are shifting - often before clear evidence exists on whether these tools improve outcomes. The hallucination problems Rowe described are not edge cases; they are routine failures that affect student assessment. Educators who understand both the capabilities and the failure modes of AI for Education tools will be better positioned to shape how their schools deploy them, rather than having deployment decisions made without their input.
Your membership also unlocks:
