Back to news
Sunday report

Sunday Signal Report: August 23, 2026

This week's thesis, under Agency and Learning: not every screen is the same job. Washington asked districts to judge tools by learning. Wichita already published the local test for families.

Rob here

Not every screen is the same job.

This week's thesis, under the primary rotating lens of Agency and Learning: the public signal is a distinction, not a product. On Aug. 20 the U.S. Department of Education issued a Dear Colleague letter on education technology and screen time. It is guidance, not a regulation. It tells states and districts to separate recreational devices from instructional tools, and to ask whether a tool solves a learning problem, for whom, for how long, and with what evidence.

That is the right fight. Schools are starting the year under statewide cellphone restrictions while classroom platforms keep adding student-facing AI. Those are not the same decision. A phone ban can protect attention. It does not design agency. A classroom chatbot can generate practice. It does not own the judgment about what a student should struggle with.

Secondary lens: Governance and Trust. In the same window the FCC's E-rate proposal is taking public comment through Oct. 13, including questions about limiting or sunsetting the program and whether funding should depend on parent opt-outs from screen-based instruction. Mixing connectivity with curriculum is how institutions lose the plot. The scarce layer is still the operating rule: named owners, exceptions, review points, and a stop condition when the tool does not help kids learn.

The local proof is already written. Under the federal letter in this report is Wichita's parent one-pager and the public family edition of our technology use guidelines: what families should see, what they should not see, and the five questions a tool has to survive.

Strong signal

Washington finally said the quiet part: judge the tool, not the screen.

What changed: on Aug. 20 the U.S. Department of Education released a Dear Colleague letter on education technology and screen time. Assistant Secretary Kirsten Baesler told states and districts to distinguish recreational technology from instructional technology, evaluate tools by learning impact rather than minutes, and expect independent evidence. The letter is guidance. It is not a new federal rule. Why it matters now: the national argument has been collapsing two different jobs into one word, screens. Phone bans, 1:1 laptops, assistive tech, and classroom AI are not interchangeable. Rob's take: this fits Agency Is Infrastructure. The five questions in the press release are the operating test: what learning problem does it solve, when, for whom, for how long, and what evidence shows it helps. Vendor logins are not that evidence. A letter that refuses to regulate can still be useful if local leaders treat those questions as procurement gates. Concrete implication for a district leader: rewrite the next edtech renewal as those five questions plus a named review owner and a dated exit test. If the vendor cannot answer them in public language, do not renew.

Local operating answer

Wichita already wrote the local test.

What changed for this report: under the Education Department letter sits the local operating answer. Wichita Public Schools already published 2025-26 technology use guidelines and a parent one-pager that separate recreational screens from instructional tools by grade band. Kindergarten has no assigned student device. Grades 1-5 share iPads. Grades 6-8 have a school laptop that stays in the building. Grades 9-12 take a school laptop home. Personal phones are a different rule. Why it matters now: the federal letter asks districts to judge a tool by the learning job, the student, the duration, and the evidence. Wichita's family documents already answer that in public language. Rob's take: this is Agency Is Infrastructure, written for parents. A phone ban can clear the desk. It does not design the lesson. The scarce layer is still the operating rule: what families should see, what they should not see, and when an IEP or 504 tool is access rather than extra screen time. Concrete implication for a district leader: put the family one-pager next to the next edtech renewal. If a vendor cannot survive the grade-band 'will not see' list, do not renew.

Assistive technology on an IEP or 504 plan is access to learning. It does not count as discretionary screen time. The plan wins.

Kindergarten

No assigned student device. The teacher uses a screen for short group lessons.

What families should see. A short video, then talk and draw. A counting song that gets kids moving. A virtual field trip followed by a hands-on activity.

What they should not see. Children sitting alone with a screen. Screens used to keep children busy or calm. Videos with no lesson afterward.

Grades 1-2

Shared iPad. Used in some lessons, not all day.

What families should see. Kids recording a story in their own words. Partners making a short digital book. A teacher-chosen practice app, then a hands-on activity.

What they should not see. Screens for most of the day. Long stretches on apps with no teacher help. Screens used as a reward.

Grades 3-5

Shared iPad. More research, writing, and projects.

What families should see. Research that becomes a project with writing, pictures, and voice. Class data turned into a chart. Lessons on being safe online and what a digital footprint is.

What they should not see. A screen most of the day. Technology used when paper would work just as well. Children using AI tools on their own.

Grades 6-8

School laptop. Stays at school. Does not go home.

What families should see. Real writing and peer editing. Source checking. Honest talk about when AI helps and when it does not.

What they should not see. Laptops open all day. Videos with nothing to do afterward. Free computer time with no learning goal.

Grades 9-12

School laptop. Goes home.

What families should see. Shared real-world projects. Podcasts, videos, or sites as part of the work. AI used as a thinking tool, not a substitute for thinking.

What they should not see. Technology replacing the teacher. AI doing the assignment. Screens all day with no analog work.

  1. What learning problem does this tool solve?
  2. For whom, when, and for how long?
  3. Does it add value that paper or talk cannot?
  4. Are students thinking, or just clicking?
  5. What happens before and after the screen, and what evidence will we review?
K-12 device strategy

The phone ban is now an operating problem, not a press conference.

What changed: the 2026-27 school year is the first full implementation window for a wave of state cellphone laws. Education Week's tracker, updated July 28, counts at least 39 states and D.C. requiring districts to ban or restrict student phones. Kansas House Bill 2299 requires schools to prohibit personal communication devices during the school day while still allowing school-issued devices. Lawrence's Aug. 6 FAQ is the local texture: store personal devices, keep IEP and medical exceptions, and do not confuse the ban with the laptop the district issued. Why it matters now: restriction, instructional design, family communication, and measured outcomes are still different jobs. A statute can clear the desk. It cannot prove students learned more. Rob's take: this fits Attention Is a Leadership Resource. Taking phones away can be the right attention design. It becomes cosplay if the same system then fills the day with low-value school screens and has no exception path parents can understand. Concrete implication for a district leader: publish a one-page matrix that separates personal devices from school-issued devices, names required exceptions, states how a family reaches a student, and lists the one or two measures you will review by winter break. Do not call the ban a learning win until those measures exist.

Policy and public systems

Do not let a screen-time fight become an E-rate sunset.

What changed: the FCC's Aug. 14 proposed rule on E-rate is now in a live comment window. Comments are due Oct. 13, with replies Nov. 12. K-12 Dive reported Aug. 18 that the notice asks whether policy changes are needed because of school screen time, and even whether the program should be limited or sunset given current connectivity. It also asks whether E-rate should depend on giving parents a chance to opt children out of screen-based instruction. Why it matters now: E-rate pays for eligible connectivity, not for classroom content and not for student laptops. Folding curriculum taste into the discount that keeps the network up is a category error with a clock on it. Rob's take: this is The Governance Gap wearing an Implementation Layer costume. You can believe phone restrictions are wise and still refuse to blow up the pipe. The honest local job is to file a comment that separates network access from instructional design, then put the screen-time rules in the device matrix, not in the funding panic. Concrete implication for a district leader: assign one owner to draft the Oct. 13 comment, put the Sept. 9 AASA town hall on the calendar, and keep the board packet from treating a proposed rule as a completed cut.

Agentic AI checkpoint

Agentic AI checkpoint: Hermes shipped stable patches; OpenClaw did not move this window

Hermes Agent published four stable tags in this window: v0.20.2 on Aug. 16, v0.20.3 on Aug. 17, v0.20.4 on Aug. 18, and v0.20.5 on Aug. 21. The latest canonical notes call v0.20.5 a patch release that rolls up merged work since v0.20.4 for Docker images, hosted deployments, and fresh installs, with full curated notes deferred to v0.21.0. That is a completed stable tag, not a new claim of production authority. OpenClaw had no new in-window release. The latest stable tag remains 2026.7.1-2 from Aug. 4. The newest prerelease remains 2026.8.1-beta.2 from Aug. 15, outside this window. Stable release, prerelease, merged work, and announcement stay separate. Concrete implication for a district leader: if an agent platform is on the table, ask which tag is stable, who owns updates, how secrets are bound to destinations, and whether restore has been tested on a fresh target. Do not treat a beta tag or a patch rollup as unsupervised operational authority.