Most K-12 teachers now say AI is undermining their students' critical thinking and their own trust in what those students know. A new study on preteens shows those habits take root years before high school. Four signals this edition demand your attention.
Today's Signals at a Glance
01Monday Classroom Signal—History/Social Studies: A federal anniversary tour just stopped in a New Mexico classroom. The model underneath it works in any district.History
02NPR and Ipsos find most K-12 teachers think AI is already eroding trust and critical thinking in their classrooms, and most schools still have no formal guidance to show for it.AI / EdTech
03EdWeek data show why teachers keep assigning homework, and the reason has nothing to do with tradition or compliance.Pedagogy
04A three-year UCSF study tracked how fast social media habits form in 9- to 11-year-olds, and what each year of increase costs them the year after.Youth Culture
Classroom Signal · Monday · History / Social Studies
History / Social Studies
A Federal Tour Stopped in a New Mexico Classroom. The Model Underneath It Travels.
On June 4, Assistant Secretary Dr. David Barker visited Ramirez Thomas Elementary School in New Mexico as part of the Department of Education's "History Rocks! Trail to Independence Tour," a national series building toward the 250th anniversary of the Declaration of Independence on July 4. Barker watched students present what they had learned about New Mexico's place in the founding story, then framed the visit around civic engagement and shared responsibility. The stop was small. The format underneath it is not.
That format is the part worth borrowing for a high school classroom: instead of a lecture about patriotism, students researched a specific local connection and explained it to an adult who came to listen. Swap "elementary students performing for an official" for "high schoolers defending a claim to a skeptical audience," and you have an assignment that does more civic work than another reading on the Continental Congress.
Try This — Ready to Use
Have each student find one document, building, marker, or local figure in your community connected to 1776, then write two paragraphs answering a harder question than "what happened here": what does this connection actually prove about how the Revolution reached, or didn't reach, the place they live? Some will find nothing — that's a real finding too, and worth defending in class.
Try This in Any Class — Today
Before students leave, ask them to write one sentence naming something from today's lesson they could explain to a younger sibling, and one sentence naming something they still couldn't. Collect both on the way out. It takes two minutes, gives you an honest read on what actually landed, and tells each student something useful about their own understanding before they walk out the door. Any subject, any grade level.
Signal Analysis
SIGNAL 01 — AI / EdTech
Most Teachers Now Say AI Is Eroding Trust in Their Classrooms. Few Schools Have a Plan for It.
The Development
An NPR/Ipsos poll of 545 K-12 teachers, fielded April 27 through May 5 and released June 5, found that 54% say AI is making it harder for students to develop critical thinking skills, 59% say it is eroding trust between students and teachers, and 57% say it is harder to know what students actually understand. Roughly three in four teachers believe AI's impact on education will outweigh that of the internet or computers. Only about a third say their school has formal guidelines for student AI use.
Why It Matters to You
That last number is the one to sit with: two out of three teachers in this poll are making AI-related calls in their classrooms with no school guidance behind them. That means every judgment call you make right now, on what counts as a violation, what counts as help, what counts as cheating, is setting precedent whether you intend it to or not. Teachers who write down their own working rules this month, even informally, will have something to point to when a parent or an administrator asks why they handled a case the way they did.
Why This Matters
A trust gap this wide does not close on its own. The teachers who name their standards now, in plain language, are the ones who will not be improvising when the first hard case lands on their desk.
Around the Corner
Expect districts to move on formal AI-use policy faster after this poll than they were planning to before it. NPR and Ipsos numbers travel through school board meetings quickly. If your building doesn't have a policy yet, this is the data point that gets one written. Better to help write it than to receive it.
Teachers Keep Assigning Homework. New Survey Data Says They're Right To.
The Development
The EdWeek Research Center's spring 2026 survey on homework found that 51% of teachers, the largest share of any answer offered, say its main value is forcing students to practice enough to actually master a concept, not testing what they already know. The finding lands in the middle of a years-long argument about whether homework helps, hurts, or simply measures a student's home environment more than their learning.
Why It Matters to You
That survey number is permission to stop relitigating whether to assign homework and start being precise about what it's for. Practice-oriented homework, low-stakes, frequent, tied to a skill students are still building, does something that a single big assessment cannot: it gives a struggling student multiple chances to get it wrong before it counts. If your homework currently functions as a check on whether students did the reading rather than a chance to build the skill, the data here points you toward a redesign that costs nothing and changes what the assignment actually does.
Why This Matters
The argument over homework has mostly been a proxy for the argument over equity and access. This data reframes it around function: practice that builds mastery is worth assigning regardless of which side of the larger debate you land on.
Around the Corner
Expect more districts to formalize "practice versus assessment" language in grading policy over the next year, separating homework completion from mastery grades. Teachers who can already explain that distinction in those terms will be ahead of the policy memo when it arrives.
A Three-Year Study Just Mapped How Fast Social Media Habits Form, and What Each Year Costs
The Development
Researchers at the University of California, San Francisco followed roughly 12,000 preteens for three years in a study published in JAMA. Daily social media use among 9- to 11-year-olds rose nearly tenfold over that span, climbing from about seven minutes a day to more than 70. Each year's jump in use predicted a corresponding rise in depressive symptoms the following year. These are children years away from your classroom, and the habit is already locking in before they arrive.
Why It Matters to You
By the time these students reach you in ninth grade, they will have spent three to five years building the exact usage pattern this study just measured rising tenfold. That changes what you're working with on day one: not a student learning to manage a new distraction, but one arriving with an established relationship to the platform and, on the data here, an established relationship to the mood dips that come with it. Knowing that doesn't change your curriculum. It changes how you read the kid who seems checked out before the bell even rings.
Why This Matters
The window for intervention on social media habits is closing well before high school. Teachers who understand what their incoming students have already been through will read disengagement more accurately than those who assume it started this year.
Around the Corner
Expect this study to surface in middle school and elementary parent communications over the next year, and expect high schools to feel the downstream effect in the form of students arriving with longer, more entrenched usage histories than previous cohorts. The earlier your feeder schools start the conversation with families, the less of this lands fully formed on your desk.
The Bottom Line — Three Things for a High-Agency Professional
1Write down your own working rules on AI use, even if your school hasn't issued any. Two sentences on what counts as help and what counts as a violation gives you something concrete to point to the first time a parent or administrator questions a call you made.
2Look at one homework assignment you give regularly and ask whether it functions as practice or as a check on compliance. If it's the second, redesign it as low-stakes repetition tied to a skill students are still building. That's the distinction the EdWeek data says actually matters.
3Assume your incoming students arrive with years of social media habits already locked in, not weeks. Read disengagement with that history in mind before assuming it started in your classroom this semester.