The evidence base

British Educational Research Association (blog) (2020). The great algorithm fiasco. BERA Blog.

Fletcher, J., et al. (2025). Automatic speech recognition performance on African American English-speaking children ages 4-9 (word error rates near 41-43%; younger children fare worse). Perspectives of the ASHA Special Interest Groups.

Hao, K. (2020). The UK exam debacle reminds us that algorithms can't fix broken systems. MIT Technology Review.

Hofmann, V., Kalluri, P. R., Jurafsky, D., & King, S. (2024). AI generates covertly racist decisions about people based on their dialect. Nature, 633, 147-154.

Sparks, S. D. (2026). AI changes its feedback on students' writing when it knows their race, gender. Education Week.

The Markup (Feathers, T.) (2021). Major universities are using race as a 'high-impact predictor' of student success (EAB Navigate risk scores; Black students labeled high-risk at 2-4x the rate of white peers; 500+ institutions). The Markup (Mar 2, 2021).

The Markup (Feathers, T.) (2021). Texas A&M drops race from student risk algorithm following Markup investigation. The Markup (Mar 30, 2021).

CSU Chico Academic Senate (2025). Resolution on artificial intelligence (excludes AI detection as 'inconsistent and may discriminate against non-native English speakers'; bars use of teaching data for LLM training without consent; 72.0% of surveyed faculty strongly agree unregulated AI risks harm). CSU Chico Academic Senate (Oct 30, 2025).

Gaines, L. V. (NPR) (2025). Teachers are using software to see if students used AI. What happens when it's wrong? (Broward's $550k Turnitin contract; a Maryland student flagged three times). NPR (Dec 16, 2025).

K-12 Dive (Merod, A.) (2023). Turnitin admits there are some cases of higher false positives in AI writing detection tool. K-12 Dive (Jun 7, 2023).

Liang, W., Yuksekgonul, M., Mao, Y., Wu, E., & Zou, J. (2023). GPT detectors are biased against non-native English writers. Patterns, 4(7), 100779.

Madden, M., Calvin, A., Hasse, A., & Lenhart, A. (Common Sense Media) (2024). The dawn of the AI era: Teens, parents, and the adoption of generative AI at home and school. Common Sense Media.

Minnesota Public Radio (2025). PhD student says University of Minnesota expelled him over an AI allegation; his SEVIS record was terminated, jeopardizing his visa status. MPR News (Jan 17, 2025).

Tech & Learning (2023). He was falsely accused of using AI. Here's what he wishes his professor did instead (UC Davis student cleared via Google Docs revision history). Tech & Learning (Dec 4, 2023).

The Register (2023). Michigan State, Northwestern, UT Austin, and others switch off Turnitin's AI detection. The Register (Sep 23, 2023).

The San Francisco Standard (Angst, M.) (2026). A Palo Alto high schooler was accused of AI cheating. His family filed a civil rights suit. SF Standard (May 11, 2026).

Turnitin (press release) (2024). Turnitin marks one year of its AI writing detector: 200+ million papers reviewed. PR Newswire (Apr 9, 2024).

Vanderbilt University Center for Teaching (2023). Guidance on AI detection and why we're disabling Turnitin's AI detector. Vanderbilt University (Aug 16, 2023).

Weber-Wulff, D., Anohina-Naumeca, A., Bjelobaba, S., Foltýnek, T., Guerrero-Dib, J., Popoola, O., Šigut, P., & Waddington, L. (2023). Testing of detection tools for AI-generated text. International Journal for Educational Integrity, 19(26).

ABC7 Eyewitness News (2024). Beverly Hills: 5 students expelled over AI-generated nude images of classmates at a K-8 school. ABC7 Los Angeles (Mar 2024).

AI Incident Database (2023). Incident 597: Students at Westfield High School shared AI-generated nude images of classmates (Oct 2023). AI Incident Database.

Associated Press (2026). AP report: Rise of deepfake cyberbullying poses a growing problem for schools. PBS NewsHour (Jan 2, 2026).

Center for Democracy & Technology (Laird, E., et al.) (2024). In deep trouble: Surfacing tech-powered sexual harassment in K-12 schools (Sep 26, 2024). CDT.

Centers for Disease Control and Prevention (2024). Youth Risk Behavior Survey: 2023 results (16% of high school students electronically bullied; ~25% among LGBTQ+ students; no equivalent national middle school measure). CDC.

Hingham Anchor (2024). Judge denies preliminary injunction in Hingham student AI case (Nov 20, 2024 ruling). Hingham Anchor.

MLA-CCCC Joint Task Force on Writing and AI (2023). Working paper: against punitive integrity enforcement and total AI bans; distinguishes surveillance from observation. MLA-CCCC.

Tech Policy Press (2025). As deepfake bans take effect, child offenders remain a stumbling block. Tech Policy Press (Jun 24, 2025).

WBUR (Jolicoeur, L. & Mullins, L.) (2024). Parents sue Hingham school officials after student punished for AI use with no written AI policy. WBUR (Oct 17, 2024).

Bastani, H., Bastani, O., Sungu, A., Ge, H., Kabakcı, Ö., & Mariman, R. (2025). Generative AI without guardrails can harm learning: Evidence from high school mathematics. Proceedings of the National Academy of Sciences.

Brookings Institution, Center for Universal Education (via NPR) (2026). Report: risks of AI in education currently outweigh benefits. NPR (Jan 14, 2026).

De Simone, M. E., Tiberti, F., Barron Rodriguez, M., Manolio, F., Mosuro, W., & Dikoru, E. (2025). From chalkboards to chatbots: Evaluating the impact of generative AI on learning outcomes in Nigeria (Policy Research Working Paper 11125). World Bank.

Suggestive World Bank working paper: rigorous RCT, but not journal peer review.

Georgia State University News (reporting a course-level randomized trial) (2022). Classroom chatbot improves student performance, study says (n=500; first-generation students' final grades 11 points higher, 75 vs 64). GSU News (Mar 21, 2022).

Gerlich, M. (2025). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. Societies, 15(1), 6.

Suggestive Correlational, cross-sectional, self-report data (n=666); cannot establish causation.

Kestin, G., Miller, K., Klales, A., Milbourne, T., & Ponti, G. (2025). AI tutoring outperforms in-class active learning: An RCT in an authentic educational setting. Scientific Reports, 15, 17458.

Khan Academy (2025). Khanmigo District Partnerships: $15/student/year, 250-license minimum. Khan Academy blog.

Kosmyna, N., Hauptmann, E., Yuan, Y. T., Situ, J., Liao, X.-H., et al. (2025). Your brain on ChatGPT: Accumulation of cognitive debt when using an AI assistant for essay writing task. arXiv preprint 2506.08872.

Suggestive Preprint, not peer-reviewed; small sample (n=54, pivotal session n=18) and methodology have drawn formal criticism. Suggestive, not settled.

Page, L. C., & Gehlbach, H. (2017). How an artificially intelligent virtual assistant helps students navigate the road to college (Georgia State's 'Pounce': summer melt down 21%, enrollment up 3.3 points, 7,000+ students). AERA Open, 3(4).

SRI International (Roschelle, J., et al.) (2016). How big is that? Reporting the effect size and cost of ASSISTments in the Maine homework efficacy study (46 schools, ~3,050 7th graders, g=0.18, p=.007). SRI International.

The Conversation (Toch, T. et al.) (2026). Despite the growth of some AI schools, research doesn't show that AI tutors are better than human teachers. The Conversation.

American Academy of Pediatrics, Center of Excellence on Social Media and Youth Mental Health (2026). Counseling patients and families on AI chatbots (younger adolescents more vulnerable to chatbot design features; companion chatbots not recommended for minors). AAP (updated Jun 2026).

American Psychological Association (2023). Health advisory on social media use in adolescence (recommends adult monitoring of use in early adolescence, ages 10-14). APA (May 2023).

American Psychological Association (2025). Health advisory: Artificial intelligence and adolescent well-being (adolescence spans roughly ages 10-25; safeguards against companion displacement of human relationships; age-appropriate defaults; likeness protection; AI literacy education). APA (Jun 2025).

CBS News (2026). Character.AI, its founders, and Google settle lawsuits over a 14-year-old's death; a 2025 ruling had treated the chatbot as a product subject to liability law. CBS News (Jan 7, 2026).

CNBC (2025). Character.AI announces removal of open-ended chat for users under 18 (Oct 29, 2025), phasing from a two-hour cap to zero, with age assurance. CNBC (Oct 29, 2025).

Common Sense Media & Stanford Brainstorm Lab (2025). Social AI companions pose unacceptable risks to children and teens under 18 and should not be used by minors. Common Sense Media risk assessment (Apr 30, 2025).

EdSurge (2023). Teens need parent permission to use ChatGPT. Could that slow its use in schools? (terms require 13+, parental permission for 13-17; OpenAI on record it does not check). EdSurge (Nov 2, 2023).

Internet Matters (2025). Me, myself & AI (UK children 9-17: 64% use chatbots; 35% say it feels like talking to a friend, rising to 50% among vulnerable children). Internet Matters (Jul 2025).

Ofcom (2026). Children and parents: media use and attitudes report 2025-26 (56% of UK children 8-17 have used AI; 11% of child AI users use it as someone to talk to or as a friend; n=3,387). Ofcom (May 21, 2026).

Office of the U.S. Surgeon General (2023). Social media and youth mental health: The U.S. Surgeon General's Advisory (up to 95% of 13-17s use social media; nearly 40% of ages 8-12 use it despite 13+ minimums). HHS, via NCBI Bookshelf (May 23, 2023).

Chalkbeat (Barnum, M.) (2026). i-Ready vs. parents and teachers: how the adaptive ed-tech tool became a villain in schools (D.C. contract challenged; LAUSD audit; Washoe renewal cut). Chalkbeat (May 19, 2026).

Chalkbeat Newark (Gómez, J.) (2024). Newark expands Khanmigo AI to 14 schools, grades 3-8, after its First Avenue elementary pilot. Chalkbeat Newark (Aug 15, 2024).

Chalkbeat Newark (Gómez, J.) (2026). Randi Weingarten's Newark AI classroom visit 'confirmed fears,' union president says, contradicting district framing. Chalkbeat Newark (Jul 7, 2026).

Evidence for ESSA (Center for Research and Reform in Education) (2025). Amira evidence review: 'Moderate' rating built on a 2013 study of the predecessor technology and one company-run comparison. Evidence for ESSA.

ExcelinEd in Action (2025). The science of reading: 15 states strengthened early-literacy policy in 2024 (40 states + D.C. with laws by year's end). ExcelinEd in Action (Jan 7, 2025).

Nickow, A., Oreopoulos, P., & Quan, V. (2020). The impressive effects of tutoring on preK-12 learning (NBER w27476: ~0.37 SD pooled; teacher- and paraprofessional-delivered tutoring strongest, especially early-grades reading). NBER Working Paper 27476.

Schwartz, S. (Education Week) (2025). AI tutors are now common in early reading instruction. Do they actually work? (no rigorous independent RCT; most students never met recommended dosage). Education Week (Nov 20, 2025).

WWNO / NPR Louisiana (Field, A.) (2024). Louisiana expands the Amira AI reading tutor to ~100,000 students with $3.6M in federal relief funds ($17 per additional student). WWNO (Oct 8, 2024).

Center for Democracy & Technology (2024). Late applications: Disproportionate effects of generative AI detectors on English learners. CDT brief.

Chalkbeat (Aldrich, M. W.) (2025). Trump administration rescinds 2015 guidance on serving English learners; underlying statutes remain. Chalkbeat (Aug 20, 2025).

EdSurge (Lieberman, M.) (2025). Schools turn to AI translation tools to support English learners (benefits and documented failure modes). EdSurge (Nov 10, 2025).

Hechinger Report (Butrymowicz, S. & Mader, J.) (2017). Lost in translation: DOJ required New Bedford schools to use trained interpreters, 'not students, relatives or Google Translate'. Hechinger Report.

National Center for Education Statistics (2024). English learners in public schools (10.6% of students; cities 9.8%, rural 4.8%). NCES Condition of Education.

Garlen, J. C. (2019). ClassDojo raises concerns about children's rights (claims use in 95% of U.S. K-8 schools; behavior scoring and data critiques). The Conversation.

Campus Technology (2022). DOJ agreement with UC Berkeley over digital content inaccessibility could have huge impact on IHEs (reliance on inaccurate YouTube auto-captions was part of the violation). Campus Technology (Dec 15, 2022).

Center for Learner Equity (2024). Students with disabilities face exclusionary discipline at 2-3 times the rate of nondisabled peers (2020-21 CRDC). CLE via ERIC (ED664042).

EdScoop (2021). Universities shift from peer note-taking to assistive technology (peer note-taking fell from 43% of accommodations in 2019 to 25% in 2021 across 95 institutions). EdScoop.

K-12 Dive (reporting CDT survey data) (2025). Nearly 60% of special education teachers used AI to help develop an IEP or 504 plan in 2024-25; 15% wrote plans in full. K-12 Dive (Nov 4, 2025).

National Center for Education Statistics (2024). Students with disabilities (7.5 million; 15% of public school enrollment). NCES Condition of Education.

U.S. Department of Education, OSERS (2024). Myths and facts surrounding assistive technology devices and services (IDEA obligations). U.S. Dept. of Education (Jan 22, 2024).

U.S. Department of Justice (2024). Final rule on the accessibility of web content and mobile apps (ADA Title II; WCAG 2.1 AA; compliance April 26, 2027 for larger public entities and April 26, 2028 for smaller ones after the April 2026 interim final rule). ADA.gov (Apr 24, 2024).

U.S. Government Accountability Office (2024). Higher education: students with disabilities (GAO-24-105614; a growing share of enrollment, driven by mental-health conditions and ADHD; self-advocacy gaps in accessing services). GAO (Apr 30, 2024).

U.S. Government Accountability Office (2018). K-12 education: Discipline disparities for Black students, boys, and students with disabilities (GAO-18-258). GAO.

The Education Trust (Patrick, K., Davis, J. C. S., & Socol, A.) (2022). Shut out: Black students are 15% of high schoolers but 9% of AP enrollees; ~225,000 Black and Latino students missing from AP. EdTrust.

Campus Technology (reporting EDUCAUSE Core Data Service 2020-21) (2022). Median central IT spending was $7.7M (4.2% of institutional spending; $1,316 per student FTE; public doctorals up to $50.5M vs community colleges up to $5.8M). Campus Technology (Mar 23, 2022).

EdBuild (2019). $23 billion: Nonwhite districts receive $23 billion less than white districts serving the same number of students ($2,226 less per pupil). EdBuild.

The Century Foundation (Baker, B. et al.) (2020). Closing America's education funding gaps (~$150B/year adequacy gap; high-Black/Latinx districts face gaps over $5,000/pupil). The Century Foundation.

American Association of Community Colleges (2026). AACC Fast Facts 2026 (10.6M students; 67% of credit students part-time; 32% first-generation; 23% report a disability; tuition $4,150 vs $11,950 at public four-years; revenue 36.2% state / 22.0% local / 20.7% tuition). AACC (Mar 2026).

American Association of State Colleges and Universities (2023). Regional comprehensive universities: the basics (517 RCUs enrolling 4.9M undergraduates; 41% Pell at AASCU institutions; more than 80% of students come from within 50 miles; 220 are minority-serving institutions). AASCU (Apr 2023).

EdWeek (reporting Why Rural Matters 2023) (2023). Report outlines the distinct challenges facing rural schools. Education Week (Nov 2023).

Federal Communications Commission (2019). Detroit's digital divide (38% of residents lack home broadband; 63% of low-income households offline). FCC blog.

Federal Communications Commission (2024). 2024 Section 706 report: 28% of rural residents and 23% on Tribal lands lack 100/20 Mbps broadband. FCC.

HBCU Money (2026). 2025 top 10 HBCU endowments (only Howard above $1B; the top-10 HBCU endowments total $2.4B vs $340B at the top-10 PWIs). HBCU Money (Feb 17, 2026).

Hechinger Report (Wright, A.) (2019). Delta teacher shortage reaches crisis levels (Delta districts 114x more likely to face shortages). Hechinger Report.

Kneebone, E. & Garr, E. (Brookings Institution) (2010). The suburbanization of poverty (suburban poor grew 25% 2000-2008, five times the city rate). Brookings.

KSDK News (2024). Ferguson-Florissant schools face $7.7 million deficit amid declining enrollment. KSDK.

National Center for Education Statistics (2026). College Navigator: University of South Carolina (38,532 enrolled; 19% Pell; 79% six-year graduation vs the 64% sector average). NCES College Navigator.

National Rural Education Association (Showalter, D. et al.) (2023). Why rural matters 2023 (7.3 million students attend rural districts). NREA.

U.S. Census Bureau (2026). Annual Survey of School System Finances FY2024 (national per-pupil spending $17,619). U.S. Census Bureau.

UNCF / PGIM (2024). PGIM-UNCF study finds private HBCU endowments need investment support (typically one investment professional vs six at comparable non-HBCU institutions). UNCF news release.

White House Council of Economic Advisers (2024). The economics of HBCUs (~75% of HBCU students rely on Pell; endowment per student roughly 50% of non-HBCU publics and 21% of non-HBCU privates; $12B land-grant funding gap; 30% of students climb two or more income quintiles). CEA (May 16, 2024).

Anthropic (2025). Anthropic Education Report: how university students use Claude (574,740 conversations analyzed; computer science 38.6% of usage vs 5.4% of degrees). Anthropic (Apr 8, 2025).

Banks, D. C. (2023). ChatGPT caught NYC schools off guard. Now, we're determined to embrace its potential. Chalkbeat New York op-ed (May 18, 2023).

Common Sense Media (2024). New report shows students are embracing AI despite lack of parent awareness (7 in 10 teens have used generative AI; 53% for homework help; 46% of school-use teens without teacher permission; only 37% of parents of AI-using teens knew). Common Sense Media press release (Sep 18, 2024).

Digital Promise (Mills, K., et al.) (2024). AI literacy: A framework to understand, evaluate, and use emerging technology. Digital Promise.

Elsen-Rooney, M. (Chalkbeat New York) (2023). NYC blocks access to ChatGPT on school networks as cheating fears swirl. Chalkbeat New York (Jan 3, 2023).

Freeman, J. (2026). Student generative AI survey 2026 (94% of undergraduates use generative AI on assessed work). Higher Education Policy Institute (Report 199).

Freeman, J. (Higher Education Policy Institute) (2025). Student generative AI survey 2025 (92% of students used AI, up from 66% in 2024; 88% used it for assessments). HEPI (Feb 26, 2025).

Google (The Keyword blog) (2025). Miami-Dade County Public Schools deploys Gemini to more than 105,000 high school students. Google blog / case study.

GovTech (reporting the EDUCAUSE 2025 AI Landscape Study) (2025). Survey: higher-ed AI adoption faces financial and policy hurdles (39% of institutions have AI acceptable-use policies, up from 23%; only 2% report new AI funding). GovTech.

Lee, I., Ali, S., Zhang, H., DiPaola, D., & Breazeal, C. (MIT RAISE) (2021). Developing middle school students' AI literacy (the DAILy curriculum: ~30 hours, piloted with 87% of participants from groups underrepresented in STEM). Proceedings of SIGCSE '21 / MIT DSpace.

Pew Research Center (2025). About a quarter of U.S. teens have used ChatGPT for schoolwork, double the share in 2023 (20% of 7th-8th graders; n=1,391). Pew Research Center (Jan 15, 2025).

UNESCO (Miao, F. & Holmes, W.) (2023). Guidance for generative AI in education and research. UNESCO.

Diliberti, M. K., Lake, R. J., & Weiner, S. R. (RAND) (2025). More districts are training teachers on artificial intelligence (RR-A956-31). RAND Corporation.

RAND Corporation (2025). Uneven adoption of artificial intelligence tools among U.S. teachers and principals (RR-A134-25). RAND Corporation.

American Council on Education (Higher Ed Today) (2023). Brief overview of U.S. faculty hiring trends (68% of faculty appointments were contingent in fall 2021). Higher Ed Today (May 1, 2023).

American Federation of Teachers (2025). AFT to launch National Academy for AI Instruction with Microsoft, OpenAI, Anthropic and the UFT ($23M over five years). AFT press release (Jul 8, 2025).

Anthropic (2025). Anthropic Education Report: how educators use Claude (57% curriculum development; grading the most automation-style use at 48.9% of grading conversations, despite being rated the task AI is least effective at). Anthropic (Aug 27, 2025).

Chalkbeat (Toulas, A. et al.) (2025). Teachers union partners with OpenAI and Microsoft on $23 million AI training academy (with expert criticism). Chalkbeat (Jul 8, 2025).

Gallup & Walton Family Foundation (2025). Three in 10 teachers use AI weekly, saving up to six weeks a year. Gallup (Jun 24, 2025; n=2,232 K-12 teachers).

Inside Higher Ed (reporting the AAUP report 'AI and the Academic Professions') (2025). Faculty often missing from university decisions on AI (90% of institutions integrating AI; 71% of faculty say administrators decide with little faculty input; 81% required to use AI-embedded ed tech). Inside Higher Ed (Jul 22, 2025).

Learning Policy Institute (2025). State teacher shortages: 2025 update (~411,500 positions unfilled or filled by underqualified teachers). Learning Policy Institute.

National Center for Education Statistics (2020). Race and ethnicity of public school teachers and their students (79% of teachers were White, 2017-18). NCES 2020-103.

National Council on Teacher Quality (2024). Paraprofessional and educator support staffing analysis (aides are the most-cited vacancy; severity highest in schools serving the most students of color). NCTQ.

National Education Association (2024). Report of the NEA Task Force on Artificial Intelligence in Education (first NEA AI policy statement). NEA (Jul 2024).

Nguyen, T. D., Lam, C. B., & Bruno, P. (2024). What do we know about the extent of teacher shortages nationwide? (~36,500 truly unfilled positions). AERA Open.

TechCrunch (Silberling, A.) (2024). Arizona's getting an online charter school taught entirely by AI (Unbound Academy, approved 4-3). TechCrunch (Dec 20, 2024).

The 74 (Peetz, C.) (2024). How districts can fund high-quality tutoring now that ESSER money is gone. The 74.

Education Week (2025). AI is changing teaching, but few labor contracts reflect it. Education Week (Aug 26, 2025).

United Teachers Los Angeles (2026). Win Our Future tentative agreement: AI cannot replace bargaining-unit positions and shall not be used to surveil employees or students. UTLA (Apr 2026).

Anderson, M. & Perrin, A. (Pew Research Center) (2018). Nearly one-in-five teens can't always finish their homework because of the digital divide. Pew Research Center.

Congressional Research Service (2024). The Affordable Connectivity Program ended May 2024, cutting subsidies for 23 million households (IF12637). CRS.

EdNC (reporting CoSN data) (2026). Districts struggle to replace student devices as ESSER-bought fleets age out with no replacement funding. EdNC (Jan 12, 2026).

Institute for Higher Education Policy (2021). Online isn't optional: student polling on access to internet and devices (28% of college students have less-than-reliable internet; Latinx students 34%, under-$50k households 31%; 8% mainly complete coursework on a phone). IHEP (May 2021; n=972).

K-12 Dive (reporting CoSN 2026 survey) (2026). Most districts still run 1:1 device programs (88-90% in secondary), but 38% of edtech leaders expect funding cuts. K-12 Dive (May 12, 2026).

Pew Research Center (2025). Internet/broadband fact sheet (54% home broadband under $30k income vs. 94% at $100k+; rural 71%). Pew Research Center.

Ballotpedia News (2025). Twenty-two states enacted K-12 cellphone bans so far in 2025 (~18 states plus D.C. bell-to-bell; Georgia and Florida apply bell-to-bell to K-8 only; Georgia's Distraction-Free Education Act effective Jan 2026). Ballotpedia (Aug 8, 2025).

Barshay, J. (Hechinger Report) (2025). Proof Points: Cellphone bans boosted scores, but suspensions of Black students rose ~16% in year one. Hechinger Report (Oct 20, 2025).

Figlio, D. & Özek, U. (2025). The effects of Florida's statewide phone ban (NBER Working Paper 34388). NBER.

Suggestive NBER working paper, not yet peer-reviewed.

Goodyear, V. A., et al. (2025). School phone policies and adolescent mental wellbeing (SMART Schools study: no difference between restrictive and permissive schools). The Lancet Regional Health - Europe.

LAist (Gilbertson, A.) (2025). LAUSD's $7 million bell-to-bell phone ban after one semester (up to 63,000 students eligible for IEP/504 exemptions). LAist.

MultiState (2026). How state lawmakers are addressing cell phones in schools (23 states enacted restrictions in 2025 alone). MultiState (Jan 15, 2026).

Intelligent.com (2023). 8 in 10 colleges will use AI in admissions by 2024 (50% already do; among users, 87% say AI sometimes or always makes the final admissions decision). Intelligent.com (Sep 6, 2023).

Suggestive Online survey of 399 education professionals commissioned by a college-rankings site. Widely reported and directionally consistent with other coverage, but not peer-reviewed research.

National Association of Student Financial Aid Administrators (AI Task Force) (2026). Use of AI in financial aid offices: survey report (n=1,233 professionals at 834 institutions; 54% used AI; only 9% aware of any institutional AI policy; 41% no review process; AI-written SAP appeals an emerging problem). NASFAA (Apr 2026).

The Stanford Daily (2023). Faculty Senate moves to allow exam proctoring, ending 102-year precedent by sidestepping student vote (honor code violations rose 136 to 191 to 391). The Stanford Daily (Apr 28, 2023).

University of Sydney (Educational Innovation) (2024). Frequently asked questions about the two-lane approach to assessment in the age of AI (secure lane plus open lane). Teaching@Sydney (Jul 2, 2024).

American Academy of Pediatrics, Council on Communications and Media (2016). Media and young minds (under 18 months: no screens except video chat; ages 2-5: at most one hour daily of high-quality, co-viewed programming). Pediatrics, 138(5).

K-12 Dive (Merod, A.), reporting RAND RR-A4412-2 (2026). 30% of public pre-K teachers used generative AI in 2024-25; screen-time concern for young children is the top stated barrier. K-12 Dive (Jan 5, 2026).

Kaufman, J. H., et al. (RAND) (2025). AI use in schools is quickly increasing but guidance lags behind (2024-25: 69% of high school, 64% of middle school, 42% of elementary teachers used generative AI). RAND RR-A4180-1.

Powell, W. & Courchesne, S. (2024). Opportunities and risks involved in using ChatGPT to create first grade science lesson plans (hallucinated a non-existent children's book; missing age-appropriate scaffolds). PLOS ONE.

Van der Weel, F. R. & Van der Meer, A. L. H. (2024). Handwriting but not typewriting leads to widespread brain connectivity: EEG evidence with implications for young children's learning. Frontiers in Psychology.

AI for Education (2025). State AI guidance for education (policy tracker). aiforeducation.io.

Biometric Update (2026). NetChoice wins permanent injunction against Louisiana's social media age-check law (Feb 12, 2026); Virginia's under-16 law enjoined Mar 3, 2026; Nebraska's enjoined Jun 27, 2026. Biometric Update (Feb 2026).

California State Legislature (2024). AB 2876: the Instructional Quality Commission shall consider incorporating media and AI literacy into math, science, and history-social science frameworks next revised after Jan 1, 2025. California Legislative Information.

Center for Democracy & Technology (2025). States focused on responsible use of AI in education during the 2025 legislative session. CDT.

Center for Democracy & Technology (2025 legislative analysis) (2025). Nevada AB 406 (2025): prohibits school counselors, psychologists, and social workers from using AI for mental-health support functions. CDT, '2025 legislative session' analysis.

Middle States Commission on Higher Education (2025). New: Use of Artificial Intelligence in Accreditation Policy and Procedures. MSCHE (Jul 2, 2025).

North Carolina Department of Public Instruction (2024). NC generative AI implementation recommendations for PK-13 public schools. NCDPI (Jan 16, 2024).

Ohio Department of Education and Workforce (2026). AI model policy for Ohio districts and schools (ORC 3301.24: every district must adopt an AI policy by July 1, 2026). Ohio DEW.

Oklahoma Legislature (2026). SB 1734, the Oklahoma Responsible Technology in Schools Act (70 O.S. section 11-120): bars AI as the primary basis for grading, discipline, placement, promotion, or retention, while stating that it does not limit district authority. Oklahoma Legislature, enrolled bill (approved May 12, 2026).

Orrick, Herrington & Sutcliffe LLP (2025). TAKE IT DOWN Act becomes law (signed May 19, 2025): federal crime for nonconsensual intimate imagery including AI-generated; 48-hour platform removal requirement; FTC enforcement; platform processes due May 19, 2026. Orrick client insight (May 2025).

Stateline (Chatlani, S.) (2026). As AI use in schools grows, lawmakers and districts scramble to set up guardrails (incl. Idaho's no-AI-teacher-replacement law). Stateline (Jun 10, 2026).

TeachAI (2025). AI guidance for schools toolkit (as of April 2025, 26 states had issued K-12 AI guidance; 18% of principals said their school or district provided AI guidance: 13% in high-poverty schools vs 25% in affluent ones). TeachAI.

The White House (2025). Advancing artificial intelligence education for American youth (Executive Order 14365). Federal Register (Apr 23, 2025).

U.S. Department of Education (2023). 2023 non-regulatory guidance: Using evidence to strengthen education investments (ESSA section 8101(21)(A) evidence levels). U.S. Department of Education (Sep 28, 2023).

U.S. Department of Education, Office of Educational Technology (2023). Artificial intelligence and the future of teaching and learning: Insights and recommendations. U.S. Dept. of Education (May 2023).

U.S. Department of Education, Office of Planning, Evaluation and Policy Development (2025). Dear Colleague Letter: Use of federal grant funds to improve education outcomes using artificial intelligence. U.S. Dept. of Education (Jul 22, 2025).

University of California AI Council (2025). UC AI governance and transparency: applicable law and policy (the system's governance-first approach to AI adoption). University of California.

9to5Google (2025). Google makes Gemini Advanced free for U.S. college students. 9to5Google (Apr 17, 2025).

Anthropic (2025). Introducing Claude for Education (Learning Mode; Northeastern's 50,000 students, faculty, and staff across 13 campuses; LSE; Champlain College). Anthropic (Apr 2, 2025).

Associated Press (2024). Founder of AI company used by schools in Los Angeles, NYC, Atlanta arrested on fraud charges (allegations). AP via WUNC (Nov 19, 2024).

Associated Press (2026). A New York district paused its $60,000 humanoid 'AI tutor' robot after community backlash. AP via NPR (Jul 29, 2026).

California State University (CSUN Newsroom) (2025). CSU announces landmark initiative to become nation's first and largest AI-empowered university system ($16.9M/18 months with OpenAI; 460,000+ students and 63,000 faculty and staff across 23 campuses). California State University (Feb 4, 2025).

Chapman, B. (LA School Report) (2024). Future of LAUSD's AI student chatbot in doubt 3 months after launch as ed-tech firm furloughs staff ($6.2M contract). LA School Report (Jun 26, 2024).

Gómez, J. (Chalkbeat Newark) (2024). Newark Public Schools wants new AI tutor after pilot testing (free pilot; ~$35/student/year at scale). Chalkbeat Newark (May 13, 2024).

GovTech (Center for Digital Education) (2024). EDUCAUSE '24: the University of Michigan's journey with generative AI (in-house U-M GPT launched Aug 2023; ~2,500 Maizey instances). GovTech.

Hickey, M. (EdTech Magazine) (2023). How institutions responded when Google ended free unlimited education storage. EdTech Magazine (Jan 31, 2023).

Local News Matters (Bay City News) (2026). CSU renews OpenAI contract at $13M/year for three years amid budget cuts; as of April 2026 only 0.7% of students and 16% of faculty had completed voluntary AI training; faculty union files unfair-labor-practice charge. Local News Matters (May 27, 2026).

MagicSchool AI (company announcement) (2025). $45M Series B; company-reported 6 million educators and 10,000+ school partnerships. MagicSchool blog (Feb 11, 2025).

Merod, A. (K-12 Dive) (2025). 3 questions for K-12 leaders to consider amid the AI tutoring boom (700,000 students; no RCT evidence). K-12 Dive (Aug 25, 2025).

Student Data Privacy Consortium (2025). National Data Privacy Agreement v2.2 (275,000+ standard DPAs executed; no AI-specific clauses). SDPC / A4L.

The Westside Gazette (2026). Broward school board approves $2M AI contract amid budget cuts (donor-funded, year-4 cost cliff flagged in public comment). The Westside Gazette (May 2026).

Young, J. R. (EdSurge) (2024). An education chatbot company collapsed. Where did the student data go?. EdSurge (Jul 15, 2024).

Associated Press (2025). Students have been arrested over AI surveillance false alarms (Tennessee 13-year-old jailed overnight; Lawrence, KS: two-thirds of alerts non-issues). AP via WUSF (Aug 7, 2025).

Associated Press (via NBC6 South Florida) (2025). Students arrested over AI surveillance false alarms (Fairview, Tennessee: 13-year-old girl jailed overnight over a flagged joke; Gaggle CEO said the school misused the system; parents sued). NBC6 South Florida (2025).

Ayer, L., Boudreaux, M., et al. (RAND) (2023). AI-based student activity monitoring for suicide risk: scant evidence of benefit, real risks of harm (RR-A2910-1). RAND Corporation.

Center for Democracy & Technology (2022). Hidden harms: Student activity monitoring targeting LGBTQ+ students. CDT brief.

Center for Democracy & Technology (Laird, E. et al.) (2022). Hidden harms: The misleading promise of monitoring students online. CDT.

Electronic Frontier Foundation (2020). ExamSoft flags one-third of California bar exam test takers for cheating (~3,190 of ~9,000 examinees). EFF Deeplinks (Dec 2020).

Federation of American Scientists (2023). Addressing the disproportionate impacts of student online activity monitoring on students with disabilities. FAS policy memo.

Hechinger Report / The Markup (Keierleber, M. & Lecher, C.) (2024). Schools were just supposed to block porn. Instead they sabotaged homework (1.9 billion blocks in one month; filters hit high-poverty schools hardest). Hechinger Report (Apr 25, 2024).

Inside Higher Ed (2021). UM-Dearborn closed the door on remote proctoring (universitywide ban; CARES funds redirected to authentic assessment redesign). Inside Higher Ed (Apr 14, 2021).

K-12 Dive (reporting CDT survey data) (2023). Teachers say student monitoring software more often flags students for discipline than safety (38% of teachers knew a student contacted by police because of monitoring). K-12 Dive (2023).

Laird, E., Dwyer, M., & Quay-de la Vallee, H. (CDT) (2025). Hand in hand: Schools' embrace of AI connected to increased risks to students. CDT (Oct 2025).

Lurye, S. (AP) & Bryan, C. (Seattle Times) (2025). Takeaways from an investigation into AI-powered school surveillance (Vancouver, WA records release; ~1,500 districts use Gaggle). AP / Seattle Times (Mar 12, 2025).

New York State Education Department (2023). Determination prohibiting facial recognition technology in New York schools. NYSED (Sep 27, 2023).

Offices of Senators Warren & Markey (2022). Investigation: EdTech student surveillance platforms need urgent federal action (none of four vendors audited for bias). U.S. Senate (Mar 30, 2022).

The 74 (citing FTC action) (2024). FTC: AI weapons-detection company Evolv misled schools about its abilities; schools allowed to cancel contracts. The 74 (Dec 2024).

The Ubyssey (2025). Proctorio lawsuit against a university learning technologist ends after five years (consent dismissal Nov 12, 2025; a 2022 injunction retained). The Ubyssey.

U.S. District Court, Northern District of Ohio (Judge Calabrese) (2022). Ogletree v. Cleveland State University (1:21-cv-00500): a remote room scan of a student's home held an unreasonable search under the Fourth Amendment. N.D. Ohio (opinion Aug 22, 2022; remedy order Dec 20, 2022).

WKTV News (2022). Utica district replaces $4 million weapons-detection system after scanner missed the knife used in a school stabbing. WKTV (Nov 2022).

Yoder-Himes, D. R., et al. (2022). Racial, skin tone, and sex disparities in automated proctoring software (Respondus Monitor: face detection 78% vs 92% by skin tone; 6.07 vs 1.19 flags per assessment; no group differences in actual cheating). Frontiers in Education.

BleepingComputer (Toulas, B.) (2025). PowerSchool breach: 62.4 million students' data stolen; ransom paid, districts extorted again months later. BleepingComputer (Sep 2025).

K-12 Dive (Schwartz, N.) (2022). Illuminate Education breach exposes 820,000+ NYC students' data, including special-education status. K-12 Dive (2022).

Office of the New York State Attorney General (2025). Attorney General James and multistate coalition secure $5.1 million from Illuminate Education (vendor promised encryption it didn't deliver). NY AG (Nov 6, 2025).

U.S. Attorney's Office, District of Massachusetts (2025). College student sentenced to four years for cyber extortion of PowerSchool and its district customers. U.S. DOJ (Oct 14, 2025).

U.S. Government Accountability Office (2022). Cybersecurity in K-12: attacks cost districts up to $1 million and weeks of lost learning (GAO-23-105480). GAO (Oct 2022).

CNBC (Field, H.), reporting FTC action (2025). FTC orders Alphabet, Meta, OpenAI, Snap, X.AI, and Character Technologies to detail AI companion chatbots' effects on children. CNBC (Sep 11, 2025).

Federal Trade Commission (2025). Children's Online Privacy Protection Rule amendments (effective June 23, 2025; compliance April 22, 2026): biometrics covered, separate consent for third-party disclosures, no indefinite retention; AI-training disclosures require consent. Federal Register, 2025-05904.

Federal Trade Commission (2023). FTC v. Edmodo: ed-tech provider unlawfully used children's data for advertising and outsourced COPPA compliance to schools ($6M penalty; algorithmic disgorgement). FTC press release (May 22, 2023).

Future of Privacy Forum (2022). The Agora finding: schools cannot require parents or students to waive FERPA rights through an ed-tech company's terms of service. FPF.

Future of Privacy Forum (2016). Guide to protecting student data under SOPIPA (California's model student-privacy law, copied across ~33 states). FPF.

Capano, G., & Howlett, M. (2020). The knowns and unknowns of policy instrument analysis: Policy tools and the current research agenda on policy mixes. SAGE Open, 10(1).

Gottfried, M. A., Stecher, B. M., Hoover, M., & Cross, A. B. (2011). Federal and state roles and capacity for improving schools. RAND Corporation, TR-989-SF.

McDonnell, L. M., & Elmore, R. F. (1987). Getting the job done: Alternative policy instruments. Educational Evaluation and Policy Analysis, 9(2), 133-152.

Schneider, A., & Ingram, H. (1990). Behavioral assumptions of policy tools. The Journal of Politics, 52(2), 510-529.

Alsharefeen, R. (2025). Faculty as street-level bureaucrats: Discretionary decision-making in the era of generative AI. Frontiers in Education, 10, 1662657.

Suggestive Peer reviewed, but a single internationalized university in the UAE: 95 faculty and 10 leaders surveyed, 17 faculty interviewed, February 2025. Read as one institution's pattern, not as a finding about faculty generally.

Berman, P., & McLaughlin, M. W. (1978). Federal programs supporting educational change, Vol. VIII: Implementing and sustaining innovations. RAND Corporation, R-1589/8-HEW.

Durlak, J. A., & DuPre, E. P. (2008). Implementation matters: A review of research on the influence of implementation on program outcomes and the factors affecting implementation. American Journal of Community Psychology, 41(3-4), 327-350.

Fixsen, D. L., Naoom, S. F., Blase, K. A., Friedman, R. M., & Wallace, F. (2005). Implementation research: A synthesis of the literature. National Implementation Research Network, University of South Florida, FMHI Publication #231.

Honig, M. I. (Ed.) (2006). New directions in education policy implementation: Confronting complexity (Ch. 1: Complexity and policy implementation). State University of New York Press.

Lipsky, M. (2010). Street-level bureaucracy: Dilemmas of the individual in public services (30th anniversary expanded ed.). Russell Sage Foundation.

Peters, S., Guccione, L., Francis, J., Best, S., Tavender, E., Curran, J., Davies, K., Rowe, S., Palmer, V. J., & Klaic, M. (2024). Evaluation of research co-design in health: A systematic overview of reviews and development of a framework. Implementation Science, 19, 63.

State Implementation and Scaling-up of Evidence-based Practices Center & National Implementation Research Network (2015). Implementation drivers: An overview. FPG Child Development Institute, University of North Carolina at Chapel Hill.

Viennet, R., & Pont, B. (2017). Education policy implementation: A literature review and proposed framework. OECD Education Working Papers No. 162.

Weatherley, R., & Lipsky, M. (1977). Street-level bureaucrats and institutional innovation: Implementing special-education reform. Harvard Educational Review, 47(2), 171-197.

Battaglia, F. (2025). Kingdon at 40: Multiple streams, multiple flaws. Journal of Policy Studies, 40(3), 25-42.

Cairney, P. (2019). Policy analysis in 750 words: Eugene Bardach's (2012) Eightfold Path. Paul Cairney: Politics and Public Policy (University of Stirling).

Suggestive A teaching post by a professor of politics, not peer-reviewed publication. Used only for the eight-step list, which is independently corroborated by Clark (2017) reading the same 4th edition.

Clark, I. (2017). Bardach's Eightfold Path to more effective problem solving. Atlas of Public Management.

Suggestive A reference-encyclopedia entry, not peer-reviewed publication. Used only for the eight-step list, which is independently corroborated by Cairney (2019) reading the same 4th edition.

Herweg, N., Zahariadis, N., & Zohlnhöfer, R. (2022). Travelling far and wide? Applying the multiple streams framework to policy-making in autocracies. Politische Vierteljahresschrift, 63(2), 203-223.

Meltsner, A. J. (1972). Political feasibility and policy analysis. Public Administration Review, 32(6), 859-867.

Pierson, P. (2000). Increasing returns, path dependence, and the study of politics. American Political Science Review, 94(2), 251-267.

Tsebelis, G. (2010). Veto player theory and policy change: An introduction. In König, Tsebelis, & Debus (Eds.), Reform Processes and Policy Change, Studies in Public Choice 16, 3-18. Springer.

Bell, D. A. (1980). Brown v. Board of Education and the interest-convergence dilemma. Harvard Law Review, 93(3), 518-533.

Bonilla-Silva, E. (2003). Racism without racists: Color-blind racism and the persistence of racial inequality in America. Rowman & Littlefield.

Bronfenbrenner, U. (1979). The ecology of human development. Harvard University Press.

Crenshaw, K. (1989). Demarginalizing the intersection of race and sex. University of Chicago Legal Forum, 1989(1).

Ladson-Billings, G. (2006). From the achievement gap to the education debt: Understanding achievement in U.S. schools. Educational Researcher, 35(7), 3-12.

Noble, S. U. (2018). Algorithms of oppression: How search engines reinforce racism. NYU Press.

Pierce, J. J., Siddiki, S., Jones, M. D., Schumacher, K., Pattison, A., & Peterson, H. (2014). Social construction and policy design: A review of past applications. The Policy Studies Journal, 42(1).

powell, j. a., Menendian, S., & Ake, W. (2019). Targeted universalism: Policy and practice. Othering & Belonging Institute, University of California, Berkeley.

Valencia, R. R. (Ed.) (1997). The evolution of deficit thinking: Educational thought and practice. Falmer Press.

Humanities and Social Sciences Communications (editorial) (2026). Retraction note: The effect of ChatGPT on students' learning performance (retracted April 22, 2026). Humanities and Social Sciences Communications.