The DOI matches its Crossref record; 33 of the 33 references it cites were found in a registry or an index; 4 of the 21 sources its claims cite were checked for support.
Policy Silence as Governance: AI Decision-Making, Responsibility, and Risk in K–12 Education
Abstract
This study examines how policy silence functions as a governance mechanism in the context of generative artificial intelligence (AI) in K–12 education. While existing research has focused on how districts regulate or integrate AI, far less attention has been given to what happens when formal guidance is delayed, incomplete, or unresolved. Drawing on a single-district qualitative case study of a Texas public school district, the analysis uses interviews and document analysis as primary data sources, while student survey data provide descriptive context regarding patterns of AI access and guidance. Findings show that policy silence actively redistributes interpretive authority to educators and school leaders, shifting responsibility for ethical and instructional decision-making onto individual classrooms without corresponding institutional support. This redistribution produces uneven enactment and may contribute to disparities in student access, guidance, and learning opportunities. Applying Critical Policy Analysis and Jencks’ framework of educational opportunity, the study shows that policy silence is not the absence of governance but a governance choice, one that shapes how access, responsibility, and fairness are determined. Equitable AI integration requires policies that pair clarity with support for professional judgment, positioning AI governance as central to contemporary school improvement.
Overview
CoChat Summary
TL;DR
The paper argues that policy silence in K–12 AI governance is not the absence of governance but a governance mechanism that redistributes interpretive authority and risk to educators and school leaders, producing uneven enactment and equity concerns. In a Texas district case with 10 interviews, documents, and descriptive survey data from 319 recent Texas high school graduates, it finds inconsistent classroom AI rules, limited school access, and unclear guidance.
Key findings
- Policy silence operated as governance: the focal district had no formal district-wide AI policy after over a year of deliberative policy development; leaders framed delay as iterative and strategic, while educators experienced it as operational ambiguity and institutional abandonment.
- Responsibility and risk were redistributed downward: teachers reported banning AI or deciding individually, with one saying, “There’s no policy right now, so it’s really up to each teacher”; another described AI as “my arch nemesis” and “sick of policing it,” while leaders used individual “I think” judgments rather than collective frameworks.
- Student survey patterns showed uneven perceived fairness and guidance: 30% said AI policies made learning less fair, 23% more fair, 21% no difference, and 26% unsure; only 37% received clear guidance, 44% unclear or inconsistent guidance, and 19% no guidance.
- Access disparities were substantial: 21% reported full AI access at school versus 82% outside school—a 61-percentage-point gap—while 41% had no school access and 7% no home access; 65% believed more AI access would provide an advantage in learning and future opportunities, 13% said no, and 22% were not sure. District context included ~15,000 students across ~25 schools, 80% Hispanic students, and ~88% economically disadvantaged.
Methods
- Qualitative single-district case study of a Texas public school district selected as a policy-in-formation context; primary case data were semi-structured interviews with 10 participants (deputy superintendent, chief technology officer, high school associate principal, middle school principal, four high school teachers, two middle school teachers) and policy documents/media sources.
- Analysis used thematic analysis with hybrid inductive-deductive coding informed by Critical Policy Analysis (Diem et al., 2014) and Jencks’ (1988) conceptions of educational opportunity; documents were analyzed through CPA stages, and interview/document findings were triangulated.
- A student survey provided descriptive context only: N = 319 first-year college students who graduated from a Texas public high school in May 2025, recruited via classroom-based convenience sampling at a Texas public university; closed-ended categorical items covered AI access, perceived fairness, and instructional guidance, reported as frequencies and percentages and not treated as case evidence.
Limitations
- Paper-stated: the study is a single-district qualitative case with a focused participant sample and is not intended to generalize across districts; the student survey draws on a statewide cohort of recent graduates, not the focal district, and serves as descriptive context; the study did not collect participants’ years of teaching experience or comparable prior AI familiarity; evaluative claims follow from the critical and moral frameworks guiding the analysis.
- Evident from the text: the survey used classroom-based convenience sampling at one Texas public university and relied on self-reported senior-year experiences, so it cannot support generalizable or causal claims; interviews were limited to 10 participants and approximately 30 minutes each, and students were not interviewed directly, leaving student perspectives available only through survey data.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
4 Valid · 1 Not valid · 13 Not checked · 115 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Under such ambiguity, however, variability in leadership capacity, knowledge, and risk tolerance can produce uneven conditions for educators and students, reinforcing differences in access, guidance, and instructional opportunity (Arar et al., 2025; Diem et al., 2014).”
What the sources say
- Not checked — no full text of the sourcecites Arar et al., 2025 source
- Partly supportedcites Diem et al., 2014 sourceExperimental — accuracy not yet measured
The Discussion (page unavailable) says education policies can advantage some groups and disadvantage others, and Engaged and activist research (page unavailable) discusses building educators’ capacity to use policy in local schools. These passages support the general concern about unequal conditions, but do not link policy ambiguity or variation in leaders’ knowledge and risk tolerance to differences in access, guidance, or instructional opportunity. (LLM, conf=0.79)
plainstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“Such non-decision-making must be understood as a consequential policy choice that shapes practice through ambiguity rather than directive control (Diem et al., 2014).”
What the sources say
- Partly supportedcites Diem et al., 2014 sourceExperimental — accuracy not yet measured
The “Power and voice in policy” section (page unavailable) treats what policy does not say and which problems remain undefined as consequential subjects for analysis. The “Policy constructions” section (page unavailable) also discusses how policy enactment affects students. Neither passage describes non-decision-making as a policy choice or says that it shapes practice through ambiguity rather than directive control. (LLM, conf=0.78)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“At the time of data collection, no formal district-wide AI policy had been adopted”
- Also stated in: § Case Selection · plain
“actively engaged in AI policy development while operating without formal district-wide guidance”
plainSingle case- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § Case Selection · plain
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“However, interviews with educators revealed that this deliberation translated into operational ambiguity for those navigating AI decisions in practice.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“AI governance operated through interpretation rather than formal directive”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
72 more claims
Create a free account to see all 133 claims and every source verdict.
- It may state two separate claims as one. 43
- Probably not a substantial claim. 18
- The quote only partly states this claim. 1
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ Policy-in-Formation Governance
“The high school principal stated, 'I have had teachers approach me and complain that the creation of the policy has taken too long.”
§ Data Sources
“Participants were purposefully selected to capture multiple perspectives on district-level AI governance and included the deputy superintendent, chief technology officer, one high school associate principal, one middle school principal, four high school teachers, and two middle school teachers (Table 1).”
Citations
Cites
Using thematic analysis in psychology
Braun, V., & Clarke, V. (2006) · Qualitative Research in Psychology · doi:10.1191/1478088706qp063oa · Cited by 194,414
Qualitative Data Analysis: An Expanded Sourcebook
(1994) · Journal of Environmental Psychology · doi:10.1016/s0272-4944(05)80231-2 · Cited by 48,420
Bachrach, P., & Baratz, M. S. (1962) · American Political Science Review · doi:10.2307/1952796 · Cited by 3,387
Algorithms of Oppression: How Search Engines Reinforce Racism
Noble, S. U. (2020) · doi:10.2307/j.ctt1pwt9w5 · Cited by 2,709
How Leadership Influences Student Learning
Louis, K. S., Leithwood, K. A., Wahlstrom, K. L., & Anderson, S. E. (2004) · Issue Lab (Candid) · Cited by 1,644
Williamson, B. (2016) · Journal of Education Policy · doi:10.1080/02680939.2015.1035758 · Cited by 522
How Leadership Influences Student Learning
Leithwood, K., Louis, K. S., Anderson, S., & Wahlstrom, K. (2004) · doi:10.59656/el-sb5787.001 · Cited by 506
Algorithms of Oppression: How Search Engines Reinforce Racism
Noble, S. U. (2020) · New York University Press eBooks · doi:10.18574/nyu/9781479833641.001.0001 · Cited by 416
The intellectual landscape of critical policy analysis
Diem, S., Young, M. D., Welton, A. D., Mansfield, K. C., & Lee, P.-L. (2014) · International Journal of Qualitative Studies in Education · doi:10.1080/09518398.2014.916007 · Cited by 337
Rethinking Case Study Research
Bartlett, L., & Vavrus, F. (2016) · Routledge eBooks · doi:10.4324/9781315674889 · Cited by 332
Research works printed but not linked: 2
- Patton M. Q. (2015). Qualitative research & evaluation methods: Integrating theory and practice (4th ed.). SAGE Publications.
- Yin R. K. (2018). Case study research and applications: Design and methods (6th ed.). SAGE Publications.
Metadata
- Authors
- Miles, A. J., & Arar, K.
- Affiliations
- Texas State University (US)
- Corresponding author
- Miles, April Joy
- Venue
- Improving Schools
- Publisher
- SAGE Publications
- Published
- 2026-08-18
- ISSN
- 1365-4802 (ISSN-L)
- 1475-7583
- OpenAlex
- W7203676062
- Semantic Scholar
- 642a0e66b333214abf745e70cf98741e5a25c65d
- Cited by
- 0
- References
- 27
- FWCI
- 0.00
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- journals.sagepub.com/page/policies/text-and-data-mining-license (tdm)
- Locations
- Improving Schools (publishedVersion)
- Topics
- Social Sciences › Education › Artificial Intelligence in Education
- Social Sciences › Political Science and International Relations › Global Educational Policies and Reforms
- Social Sciences › Education › Digital Education and Society
- Keywords
- AI governance, generative AI, K-12 education, school districts, artificial intelligence integration, teacher responsibility, interpretive authority, educational opportunities, school leadership, professional judgment
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-11, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-11 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Policy Silence as Governance: AI Decision-Making, Responsibility, and Risk in K–12 Education. https://app.cochat.ai/papers/10.1177/13654802261470692
- The paper
- Miles, A. J., & Arar, K. (2026). Policy Silence as Governance: AI Decision-Making, Responsibility, and Risk in K–12 Education. Improving Schools. https://doi.org/10.1177/13654802261470692
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.