How K-12 Teachers Adapt Generative AI to Local Classroom Demands: Cultural Distance Between GenAI Defaults and Everyday Teaching Practices

Qing Xiao; Ruiwei Xiao; Hanqi Jane Li; Phenyo Phemelo Moletsane; Xinying Hou; John Stamper; Hong Shen

cs.HCSep 13, 2025

How K-12 Teachers Adapt Generative AI to Local Classroom Demands: Cultural Distance Between GenAI Defaults and Everyday Teaching Practices

Authors: Qing Xiao, Ruiwei Xiao, Hanqi Jane Li, Phenyo Phemelo Moletsane, Xinying Hou, John Stamper, Hong Shen

Organizations: Carnegie Mellon University, Pittsburgh, Pennsylvania, USA · University of California, San Diego, La Jolla, California, USA · University of Michigan, Ann Arbor, Michigan, USA

Abstract

Generative AI (GenAI) is entering K-12 classrooms, but its default culture does not always fit local teaching. Existing accounts either treat this misfit as macro-level structural bias or examine it through single-case studies; neither captures how much work teachers must do to make outputs usable across tasks. This exploratory, theory-building qualitative study introduces cultural distance, the gap between GenAI's default cultural repertoire and the situated demands of teachers' practice. Through interviews with 30 K-12 teachers who frequently use GenAI, 10 each from South Africa, Taiwan, and the United States, we identified three levels defined by the effort teachers invested and whether it yielded a usable output: low (minor edits), mid (sustained prompting and revision), and high (structural gaps rooted in unsupported languages, undigitized knowledge, or policy filters that no teacher effort can close). We present cultural distance as a promising framework requiring further examination and draw implications for designers, policymakers, and educators.

Figures & tables

Figure 1: The Three-Phase Process Through Which the Cultural Distance Framework Was Developed

ID	Gender	Years of Experience	School Type	Teaching Subject	GenAI Tools Used	Cultural Distance Mentioned
S1	Male	0–5	Public High School	English, Sepedi (Local Language)	ChatGPT, Meta AI, Grok	H-1
S2	Female	0–5	Public High School	Math, English	ElevenLabs	L-2, M-2
S3	Female	11–15	Public High School	English, Afrikaans (Local Language)	ChatGPT, Meta AI	L-2, H-1
S4	Female	0–5	Public Elementary School, High School	Math, Science, English, Physical Education	ChatGPT	L-2, H-1
S5	Female	0–5	Public High School	Business, Life Orientation	ChatGPT, Meta AI	H-1
S6	Female	0–5	Public High School	English, Economics, Tourism	Gemini, Meta AI	L-1, M-1, H-1

Table 1: Participant Demographics

Level	Effort reported by the teacher	Outcome	Indicative markers in accounts	Forms
Low	Minor edits to wording, tone, or format; a single prompt usually sufficed	Usable output, often exceeding expectations	”I just changed a few words”; output adopted almost directly	L-1, L-2
Mid	Sustained prompting, manual revision, supplementing with own pedagogical or cultural knowledge, or switching to an education-specific tool	Usable output, but only after considerable effort	Repeated reprompting; ”I have to change the details all by my own”; use of textbook content or local references in prompts	M-1, M-2
High	Multiple strategies attempted, or no strategy available	No usable output; task abandoned or done without GenAI	Wrong or refused responses in local languages; hallucinated local knowledge; queries blocked by product filters or institutional bans; reversion to English or to non-AI methods	H-1, H-2

Table 2: Operational Criteria for the Three Levels of Cultural Distance

Figure 2: A Spectrum of Cultural Distance in Teachers’ Experiences With GenAI

Form	South Africa (n = 10)	Taiwan (n = 10)	United States (n = 10)	Total
L-1 Communicating with parents and students	3	1	6	10
L-2 Designing engaging instructional activities	7	8	5	20
M-1 Generating assessment questions	3	3	4	10
M-2 Generating culturally relevant learning activities	2	3	1	6
H-1 Local languages and traditions missing from training data	7	1	0	8
H-2 Queries blocked by policy-level controls	0	1	1	2

Table 3: Number of Teachers Whose Accounts Contributed to Each Form of Cultural Distance, by Context

Figure 3: How to Apply the Cultural Distance Framework

Mechanism	Distance level	Motivating finding	Primary responsibility
Teacher-facing localization controls	Mid	Teachers re-improvised grade level, difficulty, and cultural framing through repeated prompting and manual editing (M-1, M-2).	Developers
Community verification workflow	High	Tools hallucinated or could not represent undigitized local/Indigenous knowledge; teachers fell back on community elders (H-1).	Developers
Policy transparency & appeal mechanism	High	Educational queries were blocked without explanation or recourse, by product filters or institutional bans (H-2).	Developers & policymakers
Cultural distance audit	All levels	The recurring low–mid–high spectrum supplies the dimensions (language, curriculum, local knowledge, policy sensitivity) a pre-deployment audit should probe.	Developers & policymakers

Table 4: Four Design and Policy Mechanisms Derived From the Cultural Distance Framework, Each Anchored to the Empirical Findings That Motivate It

Generative AI in Education

Generative AI in education uses language and multimodal models to support teaching, learning, feedback, assessment, and educational design.

See all 287 papers

Rethinking Generative AI Literacy: An Integrative, Developmental, and Dialectical Framework for K-12 Teacher Education
Aug 3, 2026Shahin Hossain, Sima Ahmadi, Leqi Li +7AI in Education AI Literacy
School of Education, University of Maryland Baltimore County, Baltimore, Maryland, USA · School of Teaching, Learning, and Curriculum Studies: Educational Technology Department, Kent State University, Ohio, USA · Department of Learning and Performance Systems, Pennsylvania State University, State College, PA, USA +6
Grounding AI-in-Education Development in Teachers' Voices: Findings from a National Survey in Indonesia
Apr 2, 2026Nurul Aisyah, Muhammad Dehan Al Kautsar, Arif Hidayat +1AI in Education
Quantic School of Business and Technology, Washington, DC, USA · Mohamed bin Zayed University of Artificial Intelligence, Abu Dhabi, UAE · Indonesia University of Education, Bandung, Indonesia
Towards Synergistic Teacher-AI Interactions with Generative Artificial Intelligence
Nov 24, 2025Mutlu Cukurova, Wannapon Suraworachet, Qi Zhou +1Generative AI in Education Human-AI Interaction
UCL Knowledge Lab, Institute of Education, University College London, UK · UCL Centre for Artificial Intelligence, Department of Computer Science, University College London, UK

Aug 3, 2026J/K move · Enter open · S save

Rethinking Generative AI Literacy: An Integrative, Developmental, and Dialectical Framework for K-12 Teacher Education

Shahin Hossain, Sima Ahmadi, Leqi Li, Idowu David Awoyemi+6

School of Education, University of Maryland Baltimore County, Baltimore, Maryland, USA · School of Teaching, Learning, and Curriculum Studies: Educational Technology Department, Kent State University, Ohio, USA · Department of Learning and Performance Systems, Pennsylvania State University, State College, PA, USA+6

AI in Education AI Literacy Generative AI in Education Responsible AI

Open paper AskPDF arXiv

Generative artificial intelligence (GenAI) has entered classrooms faster than teachers have been prepared to use it well, producing a GenAI literacy lag in which technological diffusion outpaces educators' conceptual, pedagogical, and ethical readiness. Established AI literacy frameworks predate the widespread adoption of large language models and, while acknowledging ethics, position it as a discrete competency rather than a constitutive commitment, with equity and agency as supplementary design principles. Recent GenAI-specific efforts address isolated features but remain fragmented. We introduce the Responsible AI Literacy in Education (RAIL-Ed) framework, developed through a systematic review and qualitative framework analysis of 67 studies (2023-2025), grounded in critical, pragmatist, sociocultural, and human-centered traditions (Freire, Dewey, Vygotsky, Shneiderman). RAIL-Ed specifies six interdependent pillars: Technical Fluency, Critical Evaluation, Human-AI Collaboration, Contextual Awareness, Ethical Reasoning, and Empowered Agency, marked by three commitments. It is integrative: the absence of any pillar produces a characteristic pedagogical failure. It is developmental: a three-level rubric (Emerging, Competent, Advanced) specifies how each pillar matures across the K-12 teacher-preparation continuum. It is dialectical: the same generative affordance can deepen or displace learning depending on the literacy a teacher brings to it, making the cultivation of that literacy, not the adoption of the tool, the object of design. By treating ethics, equity, and agency as constitutive, RAIL-Ed offers a theoretically grounded basis for curriculum design, teacher education, and policy, aligned with the UNESCO AI Competency Framework for Teachers and the OECD/European Commission AILit Framework. The framework is conceptual, advancing falsifiable propositions for empirical validation.

Abstract

Generative AI in Education

Explore similar work

Rethinking Generative AI Literacy: An Integrative, Developmental, and Dialectical Framework for K-12 Teacher Education

Grounding AI-in-Education Development in Teachers' Voices: Findings from a National Survey in Indonesia

Towards Synergistic Teacher-AI Interactions with Generative Artificial Intelligence