Learn21's Weekly EdTech Update

August 27, 2026

Banning AI Won't Teach Kids to Use It Well. Here's What Might
By Danni Tymitz

Legislatures across the country are grappling with how to address AI in schools. In Ohio, every school district is required to have an AI use policy. That part isn't optional, and it isn't really the hard part either. The hard part is what that policy looks like in practice, in a classroom, on a Tuesday, for a specific assignment. Because right now, the people responsible for answering that question are coming at it from completely different places. Some teachers are still wary of AI and treat it like something to be kept out of the room entirely. Others have leaned in hard and built it into everything. Most are somewhere in between, doing their best to find footing on ground that's still shifting under everyone.

That's not a knock on anyone. It's just where we are. And it means the actual, felt experience of AI policy in most schools isn't one clear stance, it's a patchwork of individual judgment calls, swinging between "not in my classroom" and "use it for everything," often without much in between.

Finding real balance, task by task, is harder than picking a side. But it's also the part that really teaches students something.

The real risk is inconsistency about AI

Here's the thing: the students in our classrooms right now are going to be using AI tools for the rest of their working lives. The question isn't whether they'll use it. It's when they learn to think about how. If that thinking only happens by accident, wherever a given teacher happens to land on the spectrum this year, most students won't get much practice with it at all. The first time many of them are asked to think critically about AI use, disclosure, or over-reliance will be in a job, a relationship, or a moment where the cost of getting it wrong is a lot higher than a rubric.

That's the case for being proactive instead of reactive. Not because AI is uniquely dangerous, but because judgment is a skill, and skills have to be practiced somewhere low-stakes before they're needed somewhere high-stakes. An assignment is exactly the low-stakes place to practice "when is this appropriate, and how do I say so honestly." A job isn't.

The problem is, "practice good judgment about AI" isn't something you can just write into a policy once and call finished. It needs a shared vocabulary: some way for a teacher to say, clearly, this is what's allowed here, and this is what isn't, without writing a new explanation for every assignment. And it needs to scale down to language a third grader can understand and up to language that makes sense in a graduate seminar.

A shared language for "how much is okay here"

That's what the FAIR Scale Toolkit is for. FAIR breaks AI use into four plain levels: Forbidden, Assisted, Integrated, Recommended, so a teacher can say exactly where a given assignment falls, instead of defaulting to whichever end of the spectrum feels safest that day.

A few things it does with that:

Classroom Planner: decide the level for a whole assignment, or break it down stage by stage (brainstorming might be Integrated, final drafting might be Forbidden), and generate ready-to-use policy language you can hand straight to students.
Badge Designer: turn that decision into a visual badge you can drop right onto an assignment sheet, so the expectation is impossible to miss or misremember.
Grade-Level Language: the same four levels, reworded for whatever room you're in, from elementary students to graduate faculty.
Disclosure Templates: sentence starters and fill-in templates that walk students through saying how they used AI, which is where the real practice happens. Disclosure isn't a gotcha; it's the habit we actually want them to carry forward.

None of this requires anyone to pick a side between fear and full embrace. It just makes the expectation visible, assignment by assignment, in language that fits the level and the audience, so "using AI well" stops depending on which classroom a student happens to be in and starts being something they actually practice, with feedback, before it matters more than a grade.

That's the whole point. Not gatekeeping AI, and not waving it through either. Building the judgment now, consistently, while the room for mistakes is still generous.

The FAIR Scale Toolkit is free to use, no account or login required. Open it, pick a level, and it's ready for your next assignment.

Try it: etla.tools/ai-fair-use-toolkit

Watch it in action below

Media Literacy Minute

The Case Against the Classroom Cellphone Crackdown

A Missouri civics teacher argues that his state’s blanket cellphone ban took away the very tool his students wanted most for their civic action project, and that schools would be better served teaching kids when and how to use their phones well. This EdSurge piece makes the case that digital citizenship instruction, not outright bans, better prepares students for a world where they will keep using these devices anyway.

Building Student Digital Literacy: ADL Innovation Lab for Educators

The Anti-Defamation League’s Innovation Lab pairs classroom teachers with ready-to-use tools and lesson plans that help students spot bias and mis- and disinformation online. This ADL resource frames digital literacy as something to build into everyday instruction rather than a one-time unit, giving teachers a practical entry point for embedding these skills across subjects.

Studies Find Media Literacy Predicts Trust in Institutions and Civic Engagement

New research out of the University of Kansas found that people who feel confident evaluating what they read online, and who pass those skills on to others, report meaningfully higher trust in institutions like government and libraries. This University of Kansas study also found that how people use social media, not just how much, predicts how civically engaged they become, underscoring why media literacy instruction matters well beyond the classroom.

What Just-in-Time Looks Like: Four AI Courses

By Carly Ghantous

The case for just-in-time professional learning is easy to make in the abstract. It gets harder the moment someone asks what it looks like in practice, in October, when nobody has a spare afternoon and the initiative is already underway.

In practice it comes down to three things: short enough that a teacher can actually finish it, available in the middle of the year rather than the week before it starts, and pointed at the specific thing that teacher needs right now. Get those three right and the learning lands close enough to the work to survive contact with it.

Here is one version of that. Our professional learning team has built four asynchronous AI courses, now available to districts. Each runs 90 minutes to two hours in lessons of three to eight minutes, delivered as self-paced modules with knowledge checks, classroom scenarios, downloadable job aids, and closed-captioned video. Each ends with an artifact the participant keeps, which matters more than it sounds. A teacher who finishes holding something usable has a reason to open the next module.

Teaching Students to Validate, Verify, and Think Critically with AI prepares teachers to move students from passive consumers of AI output to critical users of it. Four modules move from how AI generates text, to verification routines students can run themselves, to bias and the "whose voice is missing" question, to academic integrity, privacy, and what we've called the new social curriculum: the judgment students need in AI-saturated spaces. Participants leave with classroom norms and activities, not talking points.

The Nine Skills of AI-Ready Classrooms helps teachers recognize nine competencies professionals use to design and manage AI workflows, and name where each one already lives in the standards they teach rather than treating it as something new to add. A full crosswalk to Math, ELA, Science, and Social Studies standards makes that case concrete. Participants leave having mapped the nine onto their own instruction, with the workforce argument in hand for conversations with students and families.

The Power of Prompting: Collaborating not Copying teaches several prompting frameworks rather than one, since different tasks call for different structures, and much of the course is learning to tell which is which. Teachers apply them to lesson design (hooks, checks for understanding, exit tickets) and to the work that quietly eats a week: family communication, rubrics, quizzes. Participants leave with a personal plan naming what they will hand to AI and how they will check it.

Using AI for Differentiation and UDL Support is built on UDL 3.0 and its emphasis on learner agency. Teachers work through representation (tiered texts, multilingual scaffolds, plain-language rewrites), engagement (interest-based hooks, choice menus, belonging), and action and expression (parallel assessment options, scaffolds for the process rather than the product), alongside a direct module on guardrails: no student information in prompts, bias review, and IEP and 504 considerations. Participants leave having run one of their own lessons through the planning workflow.

These four courses were built with intentionality: standards-based, tool-agnostic, written for any grade level or subject area with no prior AI experience assumed, and shipped as common cartridge files that import into whatever LMS you already run. Customization is available, but these are ready to run as they are.

If any of this is on your list for the coming year, email the Professional Learning team to get details on the courses and on custom work. We are happy to talk through what would actually be useful for your staff.

Ninety minutes, broken into five minute segments, in the middle of a school year, on the thing a teacher needs this month. That is the whole idea.

Big Ideas

Big Ideas This Week: The Rules Finally Show Up

Back to School, Meet Back to Policy

As students returned to classrooms this month, a wave of districts rolled out their first real AI rulebooks: not bans, but frameworks for how the tech fits into daily learning. Charleston County, SC launched a districtwide policy this year aimed at teaching students and staff how to use AI responsibly rather than trying to keep the technology out of classrooms, with the superintendent calling it one of the first comprehensive district approaches in the nation. Meanwhile, Central Florida districts adopted a patchwork of their own rules just in time for the new year. Orange County, Volusia, and Marion County each rolled out approved-tool lists and new AI coursework, with one school board member summing up the mood: "It's like building a plane while we're flying."

What to Do When AI Cannot Detect AI? Fire The AI Detectors

A growing number of universities are quietly banning AI-detection tools because they're unreliable indicators of cheating. In their place, schools are redesigning assessments themselves: George Washington University named a special advisor on AI, the University of Manchester built a four-tier system for labeling how much AI is allowed on any given assignment, and Syracuse is weaving AI fluency directly into coursework rather than leaving it to individual professors to sort out.

The Kids Didn't Wait for Permission

While districts have spent the summer drafting policy, most students already had the habit locked in. Common Sense Media's 2026 census found 86 percent of U.S. children and teens ages 9 to 17 had used or interacted with generative AI, and 24 percent did so daily. Among young people who use AI, 85 percent reported using it for schoolwork. The real gap isn't adoption, it's guidance: as one analysis put it, "the defining trend isn't that kids are using AI—it's that they're using it with almost no instruction on how to use it wisely."