AI is no longer a website a school can choose to block. It is the surface students think, write, and learn on. Tenet exists to put that surface back under the school's control, safely, privately, and in plain language.
A few years ago, governing technology in a school meant managing a list of websites. Today a student can open a chat box and get an essay, a lab report, a counselor, or a way around every rule, on any device, in seconds. The old tools were built to filter pages. They were never built to understand a conversation.
Schools are not asking to ban AI. They are asking for the same thing they have always had with every other tool: the ability to decide how it is used, see that it is working, and keep students safe while it does. That ability does not exist yet. We are building it.
Tenet's job is to sit between students and every AI they touch, and give the school three things it cannot get anywhere else.
Districts and teachers write the rules in plain English, and Tenet enforces them on every supported AI, down to a single teacher's classroom. The school decides what good use looks like, not a vendor.
On-device safety models watch for self-harm, bullying, and crisis in real time, and private student data is scrubbed before anything ever leaves the browser. Safety and privacy are not a tradeoff.
Schools finally see how AI is actually being used, which tools, which patterns, where the risk is, in aggregate and never at the cost of a student's privacy.
The browser is where school AI lives today, and it is the one place we can govern every tool at once without a slow integration for each one. So that is where we begin, and it already works. From there, the surface only grows.
One extension governs every AI platform a student opens in Chrome, on Chromebooks and managed Windows. No proxy, no slowdown, working in classrooms now.
Deeper roster sync through ClassLink and Clever, more operating systems and surfaces, and coverage that follows the student wherever AI shows up next.
The same safety engine, grown into an on-device agent that understands intent in context, on every device capable of running it.
Everything Tenet does today runs on the device, in milliseconds, without sending a student's words to a server. That architecture is not just a privacy choice. It is the runway for what comes next.
Right now our on-device models recognize patterns: a rule being broken, a crisis signal, a jailbreak attempt. As local models get capable enough to run well on the hardware schools already own, that same engine grows into a true local AI compliance agent, one that reasons about what a student is actually trying to do, in full context, and applies the school's plain-English intent rather than a keyword.
It stays on the device. The student's words still never leave. But the guardrail stops being a filter and becomes something closer to a thoughtful aide that understands the assignment, the age, and the rule, and helps keep the student inside it. That is the destination: judgment, run locally, owned by the school.
Hand a prospect this page first for the why. Then go feature by feature when they ask how each piece actually works.
See the feature guides →