K-12Issue 505 records
K–12
Providence parents told their school board they could not opt their K-5 children out of Amira, an AI reading tutor that listens to students read, and asked what happened to voice recordings made before August. Albuquerque had already had the vendor delete prior recordings, limited assessment audio to 48 hours and offered a paper alternative. Chalkbeat and the Baltimore Focal report that AI in schools is becoming a school board election issue, with concrete rules left to districts. A new reanalysis of a Nigerian randomized trial finds six weeks of teacher-guided Copilot tutoring raised scores by 0.31 SD, though AI's own contribution was not isolated.
- Evidence records
- 5
- Cross-source patterns
- 2
- Evidence classes
- 3 independent reporting1 standards or public-body guidance1 academic research
- Outcomes
- 3 emerging1 cautionary1 effective
- Source freshness
- 4 new this fortnight1 recent
- Research completed
- 2026-10-27
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
AI in schools is becoming a school board election issue while concrete rules sit with districts
Chalkbeat reports that states mostly leave concrete AI restrictions to districts and that board candidates are campaigning on technology limits; the Baltimore Focal shows Maryland's 135-plus board candidates splitting between supervised use, K-5 bans and AI-free zones under a law that leaves plans to districts.
Operating questionIf a new board changes AI rules after November, can the district show each tool's purpose and evidence and disable features by grade quickly?
Supporting evidenceChalkbeat (national reporting)The Baltimore Focal (Focal News), candidate survey reporting
Speech-recording AI reading tools force districts to define voice-data retention, notice and alternatives
Providence parents asked for notice, consent, opt-out and an accounting of Amira voice recordings made before storage was disabled in August. Albuquerque had already obtained vendor deletion of prior recordings, set 48-hour retention for assessment audio and offered a paper assessment alternative.
Operating questionFor each speech-based AI tool, what audio and derived data are kept, for how long, who was told, and what is the non-AI alternative?
Supporting evidenceProvidence Public School District (Rhode Island), as reported by Uprise RIAlbuquerque Public Schools (New Mexico)
Evidence ledger
Full record · every source keeps its link and limitations
Providence parents tell school board they could not opt children out of Amira AI reading tutor and ask what happened to voice recordings made before August
At a Providence School Board meeting, parents said they learned by word of mouth that their K-5 children use Amira, an AI tutor that listens to students read aloud, and were told they could not opt out because the district treats it as core curriculum. They asked for advance notice, consent and opt-out rules for classroom AI tools, and for an accounting of voice recordings made before the district disabled storage of new recordings in August. A PTA leader said the tool repeatedly fails to understand a child with a lisp and argued Section 504 obligations. The curriculum director said Amira is used in all K-5 schools, recommended at 30 minutes a week, that the district does not keep recordings ('set the setting to zero'), and that she would follow up with the company. The board took no vote and set no date to discuss consent rules.
Why it matters, evidence and limitations
- Why it matters
- Directly K-12: a district-wide, state-introduced AI reading tutor for young children is being challenged on notice, consent, opt-out, retention of biometric-like voice data and accuracy for children with speech differences and bilingual learners. These are the controls U.S. districts will be asked about for any speech-based AI tool.
- Evidence and measured results
- Meeting-based reporting of public comment and staff statements; no documents, data audit or contract terms were inspected. The district official cited an Amira effect size of 0.4; that is an operator/vendor claim that the article does not substantiate and this record did not verify. The parent's claim that the district committed to 30 minutes a week in exchange for free licenses was disputed by the official.
- Limitations and uncertainty
- Single meeting; parents' and officials' statements are unverified. The publisher states the story was summarized from the meeting transcript with an LLM and may contain transcription errors. Meeting date is described only as 'Wednesday'. The 0.4 effect size is unverified.
Albuquerque Public Schools has Amira delete all prior student voice recordings, sets 48-hour retention for assessment audio, and offers a paper reading assessment alternative
After family concerns, Albuquerque Public Schools says Amira deleted all APS student voice recordings from the previous school year and earlier, and the district posted a Certificate of Data Deletion. Amira Assessment recordings, used for the state-required reading assessment, will be deleted within 48 hours so teachers can re-listen and double-check results. Families may refuse the computer-based assessment and receive a paper-and-pencil reading assessment instead. Amira Tutor, one way schools meet the state's 30-minute weekly reading tutoring requirement, will not record or store students' voices, and school instructional councils may choose another approach.
Why it matters, evidence and limitations
- Why it matters
- Why now: Providence parents on October 26 asked the same questions (historical recordings, retention, opt-out) for the same product, and New York City banned Amira outright in its newly published list. Albuquerque's August letter is a concrete, already-implemented set of controls that districts facing those questions can compare against. It is older than the last run and included only as this comparison point.
- Evidence and measured results
- Official district statement describing actions taken; the deletion certificate is linked but was not opened. No audit or independent verification of deletion or retention behavior.
- Limitations and uncertainty
- District self-report; published August 26, before the last run, and included as a comparison for current events rather than as new news. Deletion certificate not inspected. No data on how many families chose the paper option or on accuracy.
Chalkbeat: Backlash to AI and screen time in schools enters midterm and school board races but has not yet split along party lines
Candidates from both parties are campaigning against AI or screen time in schools: California Republican Steve Hilton says he would ban AI in California schools on day one, Texas Democrat Gina Hinojosa opposes AI use by elementary students, and Texas Gov. Greg Abbott promised to reduce classroom screen time. At least 13 states have passed laws on AI in education and about three-quarters have issued guidance, but concrete restrictions are often left to districts; at least six states enacted screen-time laws this year. In San Francisco, all but one of eight school board candidates said they would support limiting classroom technology. A political scientist calls AI in education an 'amorphous' issue for voters, most of whom are not parents of minors.
Why it matters, evidence and limitations
- Why it matters
- District AI policies adopted this fall may be revisited after November elections; board candidates are making technology limits a platform. Vendors and districts should expect political review of AI deployments, especially for younger grades.
- Evidence and measured results
- Reporting with named candidates, a screen-time consultant, a political scientist and AFT's president; state-law counts are cited from Chalkbeat's own trackers, not independently checked here. No polling of voters is presented.
- Limitations and uncertainty
- Political reporting; candidate statements are positions, not policies. Counts of state laws come from Chalkbeat trackers not inspected here. The page title and headline wording differ; this record uses the page title.
Baltimore Focal: Maryland school board candidates favor supervised AI over bans, with some calling for K-5 bans, AI-free zones and security reviews
More than 135 candidates are running for school boards across Maryland's 24 systems as districts implement the Artificial Intelligence Ready Schools Act. In a candidate survey, many do not support outright bans but want carefully supervised use; proposals range from a K-5 ban (Cecil County candidate), grade-by-grade guidelines and 'AI-free zones' (Montgomery County candidate) to security reviews of approved tools with no personally identifiable student data (Somerset County candidate). The article describes existing district policies: Baltimore City permits AI for brainstorming, translation and tutoring and notes AI helped draft the policy; Wicomico requires staff to verify AI-generated materials before sharing; Caroline permits learning and accessibility uses while prohibiting harmful conduct.
Why it matters, evidence and limitations
- Why it matters
- Maryland is the clearest U.S. case of a state mandate leaving concrete AI rules to districts; the incoming boards will decide approved tools, training and data protections. Complements archived coverage of the October 22 policy deadline and districts that missed it.
- Evidence and measured results
- Candidate survey responses quoted selectively; the article does not state how many candidates responded or how the survey was run. District policy descriptions are summaries, not inspected policy texts.
- Limitations and uncertainty
- Survey methodology and response count not reported; quotes are selective. Deadline description differs from archived sources. Candidate views are not policy.
Education Sciences reanalysis of a Nigerian secondary-school RCT: six weeks of teacher-guided Copilot tutoring raised overall assessment scores 0.31 SD, but AI's own contribution cannot be isolated
Using the World Bank's public replication package, the authors re-estimated a trial in which 1,328 interested first-year upper-secondary students were randomized; 657 were offered twelve 90-minute after-school sessions over six weeks (June-July 2024) combining Microsoft Copilot, curriculum-aligned English activities, pair work and teacher guidance, while 671 received regular instruction. Intention-to-treat effects were 0.310 SD on the overall assessment (95% CI 0.176-0.444; n=654), 0.238 SD in English, 0.309 SD in AI knowledge and 0.206 SD on the regular third-term school English exam; digital skills (0.139 SD) was imprecise. Effects were larger for female students and higher prior achievers. The digital-access interaction was negative but imprecise (-0.182 SD). Assessments were observed for 64.2% of treated versus 50.2% of control students; inverse-probability weighting gave 0.298 SD, and excluding one large school lowered the estimate to 0.173 SD.
Why it matters, evidence and limitations
- Why it matters
- Why now: published this week and not in the archive, it is an independent reanalysis that complements archived cautionary evidence (homework-outsourcing learning penalty, sparse Khanmigo engagement). It supports a structured, teacher-guided package rather than open access. Transferability to U.S. districts is limited: Nigerian urban schools, volunteer students, after-school setting, older Copilot version and short follow-up.
- Evidence and measured results
- Randomized trial, secondary analysis; OLS with school fixed effects, baseline adjustment and HC3 errors; IPW, 2,000-permutation inference and leave-one-school-out checks. The authors corrected a coding error in the original internet-access variable. Limitations: differential attrition, no AI-free tutoring arm, no persistence data, no cost data, no multiple-comparison adjustment.
- Limitations and uncertainty
- Effect is for the complete package (extra time, curriculum, pair work, teacher guidance); AI's independent contribution is not identified. Endline observed for only 759 of 1,328 students with differential attrition. Interested volunteers, 79% female sample, six-week horizon, urban schools with labs. Secondary analysis by authors not involved in the original trial; the original World Bank working paper was not opened here.
How to read this edition
Source findings, measured results and limitations come from the cited publications. Patterns, operating questions, role takeaways and implementation considerations are Lighthouse Advisory interpretation, stated as questions to validate locally rather than guaranteed outcomes. Vendor and operator claims are labeled as claims. Full research method.
- Independent reporting
- Independent reporting with attributable sources but without a formal evaluation design.
- Standards or public-body guidance
- Normative or advisory guidance from a standards body or public institution.
- Academic research
- Research produced through an academic institution or peer-reviewed venue.