Frederick County Public Schools in Maryland voted 6-1 on Wednesday, October 7, 2026, to ban student use of artificial intelligence tools from pre-kindergarten through 12th grade, according to TechSpot and Yahoo News. The vote lands just weeks after the district let middle and high school students pilot Google Gemini with teacher permission, making it one of the sharpest classroom-AI reversals of the 2026-27 school year.
The Frederick County school board’s decision did not follow a single publicized safety incident. Instead, reporting from both outlets points to a slower build-up: board members cited cheating, overreliance on chatbots, teachers who could not disable Gemini on student devices, and a rollout that one board member said happened without warning to parents. The policy now joins a growing, uneven patchwork of district-level AI rules spreading across the United States heading into the back half of 2026, and it offers a case study in how fast a school system can go from piloting a Google product to shutting it out of the classroom entirely.
What Frederick County actually voted to ban
Frederick County Public Schools, based in Frederick, Maryland, had already blocked most generative-AI platforms on district-owned devices before the 2026-27 school year started. Google Gemini was the exception. The board had approved a pilot that let middle and high school students use Gemini for schoolwork with a teacher’s sign-off, a decision TechSpot reports was driven by Scott Murphy, the district’s director of curriculum and instruction, who said Gemini was chosen because it was the only AI tool the board evaluated that met Maryland’s standards for student safety and privacy.
That pilot lasted only a matter of weeks before the board reversed course. On October 7, the board voted 6-1 to extend the restriction across every grade level, pre-K through 12. Board members initially discussed limiting the ban to younger students, then broadened it to cover high schoolers too, according to Yahoo News. The lone dissenting vote came from a board member who warned, per Yahoo’s reporting, that restricting AI access now could leave older students less prepared for college and the workforce.
The policy is not an absolute blackout. District officials clarified it does not mean students will never touch AI in a Frederick County classroom. Carve-outs remain for lessons specifically about AI literacy and for accommodations tied to students with disabilities. Those exceptions mirror the pattern seen in other districts that have tightened AI rules this year: a near-total default ban with narrow, supervised exceptions rather than a blanket prohibition with no off-ramp.
Why the board pulled back so fast
TechSpot’s reporting captures the sharpest complaint from inside the board meeting: one member said it was frustrating that Gemini had been rolled out to students without warning to parents, pointing to concerns that AI tools have been used elsewhere to generate inappropriate images of students and to bully classmates, and had in some cases produced what the member described as “bad advice from a mental health standpoint.” That same member raised the practical problem that teachers had no way to disable Gemini on student devices even when they wanted to.
Yahoo News frames the broader rationale as weeks of school board discussion about potential student misuse, cheating, and a worry that leaning on AI chatbots would erode independent critical-thinking skills before students build them. Justin Heid, president of the Frederick County Education Association, told Yahoo that the union had not heard a flood of complaints from teachers about AI in the classroom, but added: “However, several educators said they are returning to traditional paper-and-pencil assignments instead of computer-based work because they are concerned students might use AI to complete assignments.” That shift back to handwritten work echoes a pattern teachers have leaned on since ChatGPT first went mainstream roughly three years ago, TechSpot notes.
Parents in Frederick County are split. Ikea Bowie, a parent quoted by Yahoo News, backed the ban in blunt terms: “I’m for it simply because I do think it can be a crutch. They’re not relying on AI to answer their questions, to write their papers and things like that when I think that they should use their brains.” Another parent, Eric Frierson, pushed back on how far the policy goes: “Not a total ban, but I guess it has instances where it can be beneficial, especially the world that we’re moving into — we want to be prepared,” he told Yahoo News.
Frederick County is not acting alone
Part of what makes this story newsworthy beyond Maryland is the company it keeps. TechSpot’s report notes that New York City banned generative AI for elementary and middle school students last month, and that the Los Angeles Unified School District is also temporarily blocking access to generative AI tools on student devices across every grade band. Neither of those moves generated as much attention as Frederick County’s, largely because neither followed a publicly piloted Gemini rollout that then got yanked within weeks.
Nearby, Fairfax County Public Schools in Virginia has taken an even more aggressive technical approach. According to WTOP, Fairfax already blocked chatbot apps including ChatGPT, Claude, and Gemini on district systems, and its board had separately banned generative AI outright for elementary students. The district has since rolled out a dedicated Google AI blocker on school-issued laptops, a response to a wrinkle that cuts across nearly every district trying to restrict classroom AI: schools can switch off standalone chatbot apps, but Google’s AI Overviews and AI Mode inside ordinary Google Search are harder to fence off with the same administrative controls.
Not every district is moving in the same direction. Some are leaning into Gemini rather than pulling back. Jeffco Public Schools in Colorado has made Gemini available to students in grades 6 through 12, leaving day-to-day classroom rules up to individual teachers, based on coverage tracked by education-technology blog Flint. Other districts, including Haddonfield School District in New Jersey, have landed somewhere in between — banning general AI use for younger grades while keeping a single, narrowly scoped tool such as Gemini Notebook as the one approved option. The split illustrates that there is no single national policy on classroom AI in the US. Each school board is making its own call, often on a compressed timeline and with limited data on outcomes.
Google’s side of the rollout
Google has not issued a public statement responding directly to Frederick County’s vote, and neither TechSpot nor Yahoo News reports one. But Google’s broader 2026 push into K-12 classrooms is well documented. According to GCN, Google opened a dedicated Gemini tab inside Google Classroom on August 10, 2026, making it available to K-12 students of all ages in districts that had already enabled Gemini, Gemini in Classroom, or Gemini Notebook. Google’s pitch to districts has consistently been that administrators retain control: schools can enable the tools by building, grade level, organizational unit, or group, and can turn them off the same way.
That control framing is exactly what Frederick County’s board says fell short in practice. The complaint that teachers could not disable Gemini on student devices suggests a gap between what Google’s admin tools are designed to do and what actually happened at the classroom level during the pilot. Whether that gap was a configuration issue specific to Frederick County or a broader pattern other districts are quietly running into is not established by current reporting, but it is the kind of detail that tends to surface again as more districts test Gemini for Education this school year.
What the research says about teens and AI homework
The Frederick County vote did not happen in a data vacuum. Common Sense Media’s research report, titled “The Dawn of the AI Era: Teens, Parents, and the Adoption of Generative AI at Home and School,” found that 70% of teens ages 13 to 18 reported having used at least one generative AI tool, and that 40% of teens reported using generative AI to help with school assignments, according to the Common Sense Media report. More strikingly for school boards weighing policy, the same research found that among teens who used generative AI for school assignments, 46% reported doing so without their teacher’s permission.
The report also found a governance gap that mirrors what played out in Frederick County: 37% of teens reported being unsure whether their school had established rules for AI use at all, and 27% said their school had no AI-use rules whatsoever. Put together, those numbers describe a landscape where a large share of students are already using AI for schoolwork, frequently outside whatever rules exist, in districts where more than a third of teens cannot even say for certain what the rules are. That is the backdrop against which Frederick County’s board chose to reverse a pilot rather than try to patch it.
How Frederick County’s reversal compares to peer districts
| District / state | Stance on student AI use | Scope | Reported source |
|---|---|---|---|
| Frederick County Public Schools, MD | Banned, with narrow exceptions for AI-literacy lessons and disability accommodations | Pre-K through grade 12 | TechSpot, Yahoo News |
| Fairfax County Public Schools, VA | Banned on school systems; separate ban for elementary grades; added a Google AI blocker on laptops | Elementary grades explicitly; chatbot apps blocked district-wide | WTOP |
| New York City schools | Banned generative AI | Elementary and middle school | TechSpot |
| Los Angeles Unified School District | Temporarily blocking generative AI tools on student devices | All grades, described as temporary | TechSpot |
| Jeffco Public Schools, CO | Gemini made available; rules left to individual teachers | Grades 6-12 | Flint |
| Haddonfield School District, NJ | General AI banned for younger grades; one approved tool retained for some use | Pre-K through grade 5 restricted; Gemini Notebook permitted in parts of the district | Flint |
What Common Sense Media’s survey shows about teen AI use
| Metric | Figure | Source |
|---|---|---|
| Teens ages 13-18 who have used at least one generative AI tool | 70% | Common Sense Media |
| Teens who used generative AI to help with school assignments | 40% | Common Sense Media |
| Of those, share who used AI for assignments without teacher permission | 46% | Common Sense Media |
| Teens unsure if their school has AI-use rules | 37% | Common Sense Media |
| Teens who say their school has no AI-use rules | 27% | Common Sense Media |
The historical arc: from ChatGPT panic to Gemini in Classroom
School districts have been whipsawing on generative AI policy since ChatGPT’s public debut roughly three years ago triggered an initial wave of outright bans, many of which were later walked back as districts tried to build AI-literacy curricula instead of fighting a losing battle against students using chatbots anyway. Frederick County’s own trajectory echoes that arc in miniature, just compressed into a matter of weeks instead of years: block most tools, carve out an exception for one trusted vendor, pilot it with supervision, then reverse the exception once problems surface. Google’s own push to add more than 30 AI tools to Classroom for the 2026 school year was framed as giving districts more granular control, which makes Frederick County’s complaint about being unable to disable Gemini on student devices a notable data point against that pitch.
The timing also lines up with Google’s broader Gemini release cadence this year. Google shipped Gemini 4 Argon as its flagship model update in 2026, and the company has simultaneously been expanding consumer and education-tier access while separately tightening free-tier limits elsewhere in the product line. That dual push, grow the education footprint while the flagship model gets more capable and more widely deployed, is precisely the kind of fast expansion that makes district administrators nervous about losing the ability to control what students see.
The cybersecurity and liability angle school boards are quietly weighing
Education technology has also become a bigger target and a bigger liability question in 2026. A breach tied to the Canvas learning management system exposed records tied to millions of students and thousands of schools earlier this year, a reminder that any platform handling student data, AI chatbot or otherwise, is a target. Lawmakers have also been pushing new AI liability frameworks following high-profile breaches tied to AI agents, which raises the stakes for any district that lets a third-party AI tool touch student accounts without a clear kill switch. Frederick County’s complaint that teachers could not disable Gemini on student devices is exactly the kind of operational gap that liability-focused legislation is aimed at closing.
Privacy regulators overseas have been moving in a similar direction. Earlier this year, UK regulators secured AI privacy pledges from a group of major technology firms, a sign that the pressure on AI vendors to prove they can lock down data handling in sensitive contexts, including classrooms, is not limited to the US. Frederick County’s decision to cite “the state’s high standards for safety and privacy” as the bar Gemini had to clear before the pilot even began shows that framing is already baked into how at least some US districts are evaluating AI vendors.
Market impact: what this means for Google and ed-tech vendors
A single county board in Maryland is not going to move Alphabet’s stock price, and Google’s education business is a small fraction of its overall revenue. But the reputational mechanics matter more than the dollar figure. Gemini for Education depends on district-level trust: administrators enabling the tool building-by-building, grade-by-grade, the same granular control Google has marketed as the safety net. Every high-profile reversal, especially one that plays out in local news with named parents and union officials on the record, raises the bar other superintendents will demand before signing off on their own pilots.
That dynamic also affects every other AI vendor trying to sell into K-12, not just Google, especially as Google pushes Gemini-based agents into workplace and productivity tools well beyond the classroom. ChatGPT, Claude, and Microsoft Copilot are all name-checked in Fairfax County’s existing block list alongside Gemini, according to WTOP, which means the entire category is being evaluated through the same skeptical lens district by district. Vendors that can demonstrate a working, teacher-controlled kill switch, not just a policy promise, are likely to have an easier time getting through procurement in late 2026 than vendors asking districts to trust an admin panel that reportedly did not work as advertised in at least one real deployment.
What happens next in Frederick County
Current reporting does not specify an exact enforcement start date beyond the October 7 vote itself, nor does it detail the technical steps the district will take to fully disable Gemini and other AI tools across every grade and device. The policy’s carve-outs, for AI-literacy instruction and for students with documented disability accommodations, suggest the district is trying to avoid an all-or-nothing approach even as it tightens the default rule. How those exceptions get administered in practice, and whether Frederick County circles back to a narrower, better-controlled pilot later in the school year, is the open question heading into the rest of 2026.
Predictions: where classroom AI policy goes from here
- More districts will follow Frederick County’s pattern of a short, supervised pilot followed by a fast reversal once a visible problem (cheating, a parent complaint, a disabling-access failure) surfaces, rather than multi-year rollouts.
- Google and other vendors will face growing pressure to ship a verifiable, teacher-level “hard off” switch for AI tools on managed devices, not just an admin console toggle, after Frederick County’s complaint that Gemini could not be disabled.
- Expect more districts to specifically call out Google Search’s AI Overviews and AI Mode as a separate enforcement problem from standalone chatbots, following Fairfax County’s dedicated AI blocker rollout.
- State-level guidance on classroom AI is likely to become more common as individual school boards signal they want a safety and privacy bar before approving any vendor, echoing the standard Frederick County initially applied to Gemini.
- The gap between districts that ban AI outright (Frederick County, parts of New York City, Los Angeles) and districts that lean in with teacher-level discretion (Jeffco) will likely widen before it narrows, since there is still no federal standard forcing convergence.
FAQ: Frederick County’s AI ban
Which district banned AI, and when did the vote happen?
Frederick County Public Schools in Maryland. The school board voted 6-1 on Wednesday, October 7, 2026, according to TechSpot and Yahoo News.
What grades does the ban cover?
The policy covers students from pre-kindergarten through 12th grade, with exceptions for lessons about AI itself and for accommodations tied to students with disabilities.
Why did the board reverse the Gemini pilot so quickly?
Reporting points to concerns about cheating, overreliance on AI, a lack of parent notice before the pilot began, teachers being unable to disable Gemini on student devices, and broader worries about AI’s impact on children, rather than one single publicized safety incident.
Did Google respond to the ban?
No public statement responding directly to Frederick County’s vote has been reported by TechSpot or Yahoo News as of this writing.
Are other districts doing the same thing?
Yes. TechSpot reports that New York City banned generative AI for elementary and middle school students last month, and that Los Angeles Unified is temporarily blocking generative AI tools on student devices across all grades. Fairfax County Public Schools in Virginia has gone further with a dedicated Google AI blocker on school laptops, per WTOP.
Is any district doing the opposite and expanding AI access?
Yes. Jeffco Public Schools in Colorado has made Gemini available to students in grades 6 through 12 and left classroom-level rules to individual teachers, according to coverage tracked by Flint.
How many teens are already using AI for schoolwork regardless of district policy?
Common Sense Media’s research found that 40% of teens ages 13-18 reported using generative AI to help with school assignments, and that 46% of those teens did so without their teacher’s permission.
Does the ban apply to students’ personal devices outside school?
Current reporting describes restrictions on student use during school and on district-managed devices; it does not establish that the policy extends to personal devices used outside school hours.
