A district or a teacher writes how AI should behave in normal sentences. Tenet turns those sentences into enforced behavior on every supported AI platform, invisibly to the student, and keeps them in force for the whole conversation.
Most AI controls give you a blunt switch: allow or block. Tenet gives you a dial you set with words. You write something a colleague would understand, for example “Be a Socratic tutor. Give hints, not answers. Never write the essay.” and Tenet makes the AI actually behave that way. There is no policy language to learn, no scripting, and no separate setup for each AI tool.
A district can allow ChatGPT, but allowing it does nothing to stop it from writing a student's essay outright. Blocking it entirely just sends students to their phones. The real need is in the middle: let students use AI, but on the school's terms. Until now that meant either trusting every student to self-police, or asking IT to build brittle, per-platform rules that break the moment a vendor changes its interface.
Plain-English rules collapse that work to a sentence. And because each teacher can write their own, the same district can run a strict exam-week classroom and an exploratory research classroom at the same time, without a meeting, a ticket, or a policy rewrite.
The same rule works whether the student opens ChatGPT, Claude, or Gemini. Tenet has a purpose-built adapter for each supported platform, so one rule set covers all of them.
The district writes guardrails every student gets, on every platform. This is included free in Basic.
With Pro, each teacher writes rules for their own class. Teacher rules are layered on top of the district baseline. A teacher can be stricter, but cannot remove a district guardrail.
Tenet recognizes the subject of a class and can tailor instructions accordingly, and supports grade-level wording so a third-grade prompt reads differently from a college-prep one.
Rules are re-applied periodically during a long conversation, so the AI does not quietly slide back into doing the work a few messages later.
District baseline (Basic): “Allow ChatGPT and Gemini for research. Never let the AI write a full essay. Always tell students to cite their sources.”
Mr. Rivera, English 10 (Pro): “Be a Socratic tutor. Ask what they have tried first. Give hints, not answers. Do not write paragraphs for them.”
Ms. Chen, Chemistry (Pro): “Help with concepts and practice problems. Do not give final answers to graded lab questions.”
Mr. Rivera and Ms. Chen never talk to IT. Each writes a few sentences in their dashboard, and their class gets exactly the AI behavior they want. That is the teacher-by-teacher control no filter offers.
In the district dashboard, write the district-wide rules and choose which AI platforms are approved. This is the floor for everyone and takes a few minutes.
Open the class, write the rules like a short lesson plan (what the AI should know, the rules it must follow, what it should do), and save. A handy frame is Information, Rules, Actions. No training required.
Pro also adds class modes for instant control without rewriting prompts: Standard (your rules apply), Ultimate Focus (only whitelisted sites, good for test prep), and Block All AI (for exams and in-class writing). Plus bell-schedule awareness, so rules can change automatically by period, and take-home rules by subject.
This split is also the upsell. A district writes its rules for free, teachers see the dashboard, and the first teacher who wants rules for their own class is the trigger to talk about Pro.
Pedagogy becomes enforceable. The AI coaches instead of answering, and you can track teacher-reported reductions in answer-dumping.
You define learning quality in your own words. It is the difference between AI gave the answer and AI coached the thinking, and setup takes five minutes.
Classroom expectations become explicit and visible instead of a hope that students behave.
You can see which classrooms have set up their rules and which have not, and nudge consistency across the building.