Culturally and linguistically adaptive AI and professional capital among language educators

Purpose This study explores how culturally and linguistically adaptive artificial intelligence (AI) systems align with patterns of professional capital—human, social, and decisional—among language educators in under-resourced, multilingual contexts. It examines how educators' engagement with the AI platform relates to knowledge development, collaborative networks, and instructional decision-making while considering infrastructural constraints and equity challenges. Design/methodology/approach A qualitatively driven multiple-case study involving 48 language educators (16 per region) from Iran, India, and rural Kenya/Tanzania was conducted. Data from semi-structured interviews, focus groups, and AI platform logs documenting feedback, collaboration and analytics were analyzed using reflexive thematic analysis in NVivo. AI platform log data contextualized qualitative interpretations rather than testing statistical relationships. Triangulation, member checking and reflexive journaling supported the credibility and trustworthiness of the findings. Findings Educators' engagement with the AI platform aligned with patterns of professional capital across three dimensions: human capital through context-aware pedagogical feedback (85% engaged with feedback features), social capital through peer collaboration (75% engaged with peer-feedback features) and decisional capital through access to analytic summaries (92% accessed analytics features). Qualitative findings further suggested that professional capital emerged through the interaction of adaptive AI, sociocultural context, and infrastructural conditions. Originality/value This study contributes empirical insights into the application of professional capital theory within AI-supported professional learning in multilingual and under-resourced educational settings. By highlighting the interaction between adaptive AI design and sociocultural contexts, the study advances understanding of culturally responsive AI integration and offers practical implications for developing more equitable, resilient, and context-responsive learning environments.