What was checked, and what changed?

How updates work

Daily review is scheduled for 6 a.m. Eastern, starting October 6, 2026. Active proposals and approaching implementation dates receive daily attention. The wider source register rotates across seven days. Substantive changes require source review before a new release; retrieval failures and unfinished checks remain visible here.

The maintenance register covers 269 source URLs from the stored inventory. Source monitoring does not establish complete national coverage. 2 older research records still await substantive verification and stay outside reviewed policy counts.

55 of 816 state/topic cells now have documented, bounded claim reviews. These are partial reviews, not completed statewide audits.

Oct 5, 2026 · Development review

Corrected grade filtering, upcoming-policy omissions and current downloads

Reviewed the educator journey from map selection to policy detail and download. Compared the reference research page. Reopened selected official grade/timing sources; attempted unresolved primary authorities. This was not an exhaustive national audit or browser session.

• Restored grade-filter metadata for 23 instructional-restriction packages using their documented scope. Previously empty grade groups hid them from grade-specific results. Remaining unclassified grades are explicitly disclosed; no grade range was inferred from age alone. National all-grade totals do not change.
• Added partial restrictions and standards-development duties to the upcoming implementation view. LAUSD future phases and Maryland AI standards now appear. Binding/guidance labels remain visible.
• Replaced stale current-download links with live district and instructional-restriction exports. Dated files remain unchanged and available in the release archive.
• Moved the map and state controls ahead of the national totals so readers can start exploring sooner, especially on smaller screens. Policy-watch filters now persist in shareable URLs and browser history.
• Reopened Arizona law, Alabama guidance, Minnesota current and historical statutory text, and Seattle’s implementation announcement. Disclosed Seattle’s January 1/January 4 internal source conflict while preserving the letter/FAQ date and Upcoming phase.
• Added regression checks for particular grade-scoped policies, omitted future measures and current exports. Map/download consistency alone had failed to catch a shared omission.
• Reopened four additional district AI pages. Philadelphia and Prince George’s retain approved-tool conditions; Long Beach retains age-specific conditions. No new grade range or implementation claim was inferred from these pages. Virginia Beach retrieval was not a completed substantive review.

Still unresolved

• Mississippi bill history returned 502; Sumner policy and Vermont pages returned 403, and the Vermont PDF was unavailable. Indexed snippets were not substituted for a current primary-body review.
• Thirteen restriction packages still lack confirmed grade-band classification. They remain available in all-grade views; grade-filtered counts are not complete national totals.
• Browser, phone-size interaction and print-preview testing remain blocked by the unavailable browser-testing capability required by the Sites skill. No visual QA or independent human review is claimed.
• State/topic completeness, district histories and delivered instruction remain open. This release corrects defects but does not establish the requested fully audited broad-launch standard.

Same-agent source and implementation review. The source-maintenance release immediately preceding this review was successfully published as version 37; ongoing monitoring does not guarantee detection of every change.

Oct 5, 2026 · Scheduled review

Source repairs and bounded maintenance review completed

Reopened all 11 active-watch source pages and the 42-URL Monday/daily queue: 36 queue URLs returned content and six failed retrieval. Availability checks are not substantive claim reviews. Additional reads covered LAUSD phases, Ohio law and course access, New Hampshire device law and Wyoming CS implementation. Four more state/topic cells received documented bounded reviews.

• Replaced the inaccessible source for five LAUSD implementation records with working district-hosted guidelines and linked the full approved policy. Confirmed grades 2–5 begin in November 2026 and grades 6–12 in January 2027; no new adoption or count change.
• Added Ohio CS Promise funding and availability caveats to the policy record and learning-support view: students are not charged, but state funding is not allocated and schools arrange provider payment.
• Reconfirmed New Hampshire’s bell-to-bell policy requirement, Ohio’s local AI-policy duty and Wyoming’s CS standards implementation memo. These checks do not establish local compliance or classroom delivery.
• Rechecked watchlist status passages: no new enactment was verified. Read Pennsylvania printer number 1697 and added its direct text link. Ohio’s October 9 amendment remains Upcoming; Fairfax’s October 21, October 31 and November 12 milestones remain meeting/recommendation dates, not confirmed policy votes.
• Preserved earlier access results in the dated maintenance artifact and updated the current source-access export. New Hampshire’s statute and Maryland’s AI-law PDF returned content on this attempt; the federal sponsor PDF failed this time, without invalidating its earlier review.

Still unresolved

• Six rotating-queue retrieval failures: LAUSD 2024 AI bulletin, Arkansas accreditation PDF, Mississippi CS agency page, Montana CS standards PDF, Alaska HB 57 text and the federal UNPLUGGED sponsor PDF. Previous substantive dates retained for failed sources.
• Portland draft attachment returned only a viewer shell. Kansas draft attachment and Massachusetts House-amendment link could not be reopened; their status pages were readable, but those texts were not freshly reverified. The UNPLUGGED Congressional identifier remains unresolved.
• Mississippi HB 702 official final status and Sumner County’s current full AI policy remain pending. Indexed policy text is a research lead, not a current-body verification. Vermont CS authority, wider state/topic coverage and district histories remain incomplete.
• No new national policy-growth claim, classroom-delivery finding, independent review or desktop/mobile/print browser audit is established by this maintenance pass.

This entry records a bounded maintenance review. The public release remains a research preview; its release date does not mean every record was reverified.

Oct 5, 2026 · Development review

Educator journey, count definitions and source-access audit reconciled

Checked availability of all 262 URLs in the starting registry. Added seven newly checked source URLs; the ledger records 269 attempts, of which 229 returned content and 40 require follow-up (including two indexed-text-only NCES sources). This is availability work, not 269 substantive claim reviews. Bounded primary reads covered Ohio, Rhode Island, Oregon, Talladega, Pew teen attitudes and a federal proposal.

• Replaced the default single-topic map with a combined limits-and-required-learning view. Added grade filters, consistent state/grade/timing links, state planning questions, source caveats, citation and print controls.
• Centralized measure definitions: personal-device rules versus instructional technology; 14 CS-instruction states versus 13 graduation-content states; 25 district restriction findings versus 34 districts in the wider collection. These are different measures, not conflicting totals.
• Reopened Ohio’s current and October 9 future statutes. Recorded the upcoming posting/date wording amendment without calling it a new phone ban. Upgraded Rhode Island to a primary-text finding.
• Corrected Oregon plan publication versus adoption, and classified its revision changelog as draft standards rather than an adopted standards-development duty. Formal adoption remains unresolved.
• Read Talladega’s district AI document through Google Docs; retained the distinction between documented teaching commitments and delivered instruction.
• Added the October 1 UNPLUGGED sponsor announcement and text to the federal proposal watch. The Congressional identifier and committee status remain unverified; it is not an enacted national ban.
• Added NCES adoption context with unlike measures kept separate. Added Pew teen expectations about AI’s personal impact, with fieldwork, population and question limits.
• Preserved dated source packages and linked a current source-access CSV, survey export and full JSON. The enabled daily maintenance workflow now includes these new sources and an immutable monthly-edition generator.
• Passed source/reference, learning, personal-device, national/district, history, graduation and release-wiring checks. Server-rendered the homepage and reconciled 48 topic/timing/grade map-download combinations. These checks are not browser or independent claim review.

Still unresolved

• National topic, proposal, district and historical coverage is incomplete. There are 51 documented bounded cell reviews out of 816 cells; no certification of exhaustive absence.
• Forty recorded URL attempts still need access follow-up. Some were blocked, oversized or returned only indexed text; they are not all dead links. The two pending research records remain excluded from reviewed totals.
• The NCES context uses readable official indexed passages; full-page retrieval timed out. Sources, measure definitions and retrieval limitations are exported.
• No independent review, educator usability study or browser/print-preview audit was completed. The required managed browser skill was unavailable.
• The daily automation is enabled for 6 a.m. Eastern. A trigger was recorded earlier, but no completed unattended publication is established by the current run log.

This manual review documents its actual scope. The public release remains a research preview; the release date is not a claim that all sources were substantively rechecked.

Oct 5, 2026 · Development review

Oregon and D.C. learning evidence narrowed; national audit remains open

Follow-up primary-source review of the three unresolved state/D.C. CS jurisdictions from the bounded learning audit.

• Added Oregon’s official statewide implementation plan as a current learning recommendation with a 2027–28 access goal; it is not treated as a universal course mandate.
• Added Oregon’s optional standards draft as Proposed and excluded it from adopted-standard counts until production-source adoption is confirmed.
• Added D.C. OSSE landing-page support evidence for Technology and Computer Science Guidance and professional learning; the linked student-facing guidance remains unavailable and no mandate is inferred.
• Reviewed CS evidence now appears in 50 of 51 state/D.C. profiles; Vermont remains the one unresolved source/adoption path in this bounded set.

Still unresolved

• Vermont agency page could not be fully retrieved; H.318 status does not establish an enacted CS requirement. No policy-absence finding.
• Oregon final optional-standards adoption status remains unresolved; the draft and expired feedback window are not presented as current law.
• D.C. linked student-facing CS guidance returned unavailable; statewide/DCPS authority and classroom delivery remain open.
• Complete state-by-topic, district census, historical and independent reviews remain unfinished. Visual browser QA is still blocked by the unavailable Sites control-browser skill.

This is a bounded manual follow-up included in the dated release; it is not a certification of national completeness or an unattended scheduled run.

Oct 5, 2026 · Development review

Learning coverage expanded; national audit remains open

Bounded review of the 12 state/D.C. jurisdictions that had no reviewed statewide CS entry. Nine yielded verified additions; three retain specific source or adoption-status gaps.

• Added 15 primary-source learning records across nine jurisdictions; reviewed CS evidence now appears in 48 of 51 state/D.C. profiles.
• Added Hawaii school-level access duties, Kansas course-or-plan duty and Missouri high-school offering requirement; preserved the difference from student graduation requirements.
• Added CS standards and integrated-standards findings, including explicit AI content in Alaska and South Dakota and AI learning recommendations in Maine and Missouri.
• Made the 12 scoped review outcomes visible in the coverage ledger and current downloads.
• Extended source monitoring to the new authorities and unresolved learning-source paths.

Still unresolved

• Oregon final optional-standards status; Vermont full agency source and H.318 disposition; DC linked guidance retrieval.
• Mississippi HB702 official final history remains unavailable; Sumner County’s full AI policy text remains unavailable. Those older records stay outside reviewed counts.
• Nationwide state/topic and district historical audits, delivery evidence and independent review remain unfinished.
• Visual browser QA is blocked because the Sites-required control-browser skill is unavailable. Source-render and consistency checks do not replace visual QA.

Included in this versioned learning-data release; these entries do not represent an unattended scheduled run.

Oct 5, 2026 · Development review

22 pending records resolved; comparisons reconciled

Revisited all 24 older pending records; substantively resolved 22 and retained two outside reviewed totals. Added two district AI-learning duties, recovered Indiana and Idaho successor rules, and added Pennsylvania’s proposal. Updated seven phone-comparison profiles and rechecked Massachusetts’s official history.

• Corrected legal scope, school implementation timing and historical status, including Alaska, Illinois, Indiana, Michigan funding, Utah council duties and Montana standards.
• Personal-device comparison now has 32 primary-text findings, 17 official-summary findings and two unresolved statewide classifications. Proposal classifications do not establish absence of other rules.
• Added Wichita and Talladega AI-learning commitments with explicit limits on what they establish about classroom delivery.
• Recovered Sumner’s October 2024 policy adoption record, removed unsupported duty claims and retained full-text review as pending.
• Added Massachusetts and Pennsylvania to the selected proposal watch; future implementation now draws from all reviewed upcoming restriction records.
• Reconciled current phone comparison, analysis, map and exports. Dated archives remain unchanged. Learning counts exclude adopted standards whose implementation timing remains unresolved.
• Expanded scheduled source coverage to include the separate device audit and learning-support registers.

Still unresolved

• Mississippi HB 702: introduced text reviewed, but official final-history retrieval remains blocked. Sumner County: adoption verified, but complete policy duties remain unverified.
• Mississippi and South Dakota statewide phone classifications remain unresolved. A failed or pending proposal does not establish policy absence.
• Some primary instruments were read from public mirrors or indexed official sections; individual records disclose those methods. No independent review is claimed.
• Complete state-by-topic, district census and matched October 2024 historical audits remain open. Published commitments do not establish delivery in every classroom.
• Visual browser QA remains blocked because the Sites preview workflow requires the unavailable control-browser skill. Source, type, rendering and consistency checks do not substitute for visual inspection.

This is a bounded claim-resolution review. A release date does not mean every source was reviewed that day, and it is not a certification of national completeness.

Oct 5, 2026 · Development review

Partial review completed

All eight watchlist sources reopened; their documented stages remain unchanged. Selected homepage survey and policy sources rechecked. Inventory-wide structural checks passed.

• Rebuilt the educator journey and repaired topic selection and shared links.
• Separated the unfinished October-to-October historical comparison from the older December-baseline artifact.
• Expanded maintenance coverage from 11 discovery URLs to a registry generated from the full stored inventory.
• Corrected a pre-existing consistency check: the full coverage index includes future, proposed and historical records; timing filters apply in the shared public model.
• Replaced stale current-download links with live exports from the shared inventory; preserved the older 230-record file as a dated archive.
• Verified the rendered homepage structure and matching map/download records across nine topic/timing views.

Still unresolved

• Complete state-by-topic and district historical audits remain open.
• Chicago AI guidebook, Alabama course-of-study PDF and Montana standards PDF could not be reopened through the web tool in this review; no new verification date assigned to those claims.
• Independent substantive review remains outstanding. Visual/browser verification was unavailable; preview requests returned an infrastructure error.
• Texas statutory pages returned their navigation shell and the Utah community-council statute returned a retrieval error; search excerpts were insufficient for a new substantive verification date.

This log records the review scope; the release is not a certification of every source or claim.