Abstract
This study examines the incorporation of Artificial Intelligence (AI) tools in bilingual primary school classrooms in the Democratic Republic of Congo (DRC), particularly emphasising instructors' interactions with AI to enhance language acquisition. The research utilised a convergent parallel mixed-methods approach, gathering quantitative data via surveys and qualitative data through semi-structured interviews with ten bilingual teachers. The principal aim was to investigate the utilisation of AI tools by teachers in bilingual environments, while the secondary objectives included identifying the hurdles encountered prior to the use of AI and exploring their perspectives regarding AI's influence on language development. The qualitative investigation, employing thematic coding, identified six primary themes: Challenges in Pre-AI Bilingual Language Acquisition, Experiences with AI Tools, Perceived Benefits, Barriers to AI Integration, Cultural and Linguistic Relevance, and Training and Support Requirements. Before the integration of AI, teachers identified substantial challenges, such as variable student language competency, insufficient multilingual resources, limited exposure to actual language usage, and inflexible educational methods. AI applications like Duolingo, ChatGPT, and Google Translate were often recognised as pivotal in altering classroom dynamics, providing individualised learning experiences, immediate feedback, and increased student engagement. Nonetheless, infrastructural impediments, specifically inadequate internet connectivity, recurrent power interruptions, and a scarcity of culturally relevant content, continued to pose significant obstacles. The quantitative results supported the qualitative observations, with a robust agreement among participants on the educational benefits of AI tools in improving student motivation, confidence, and language ability. The lack of formal training was identified as a significant deficiency, with participants calling for practical, hands-on professional development suited to their language and infrastructural circumstances. The study finds that AI can greatly improve bilingual language acquisition in under-resourced educational environments, but its practical application relies on culturally relevant content, dependable infrastructure, and ongoing professional support. These
iv
findings offer practical guidance for policymakers, developers, and educators seeking to promote inclusive and effective AI integration in multilingual classrooms.