The Role of Educational Chatbots in Adapting Educational Systems to the Bani Model: a Causal Layered Analysis Approach

This paper analyzes the integration of educational chatbots within contemporary education systems shaped by volatility, complexity, and uncertainty, using the BANI model (Brittle, Anxious, Non-linear, Incomprehensible) as a conceptual lens. The study employs Causal Layered Analysis (CLA) to investigate the symbolic, systemic, and functional significance of conversational technologies in education. A theoretical architecture for a retrieval-augmented generation (RAG) chatbot is proposed, designed to process institutional educational documents and deliver context-aware responses based on semantic matching. The architecture is detailed with a focus on document processing workflows, vector-based indexing, and natural language generation. Potential applications are discussed in higher education with emphasis on benefits such as cognitive support, personalization, reduced information anxiety, and institutional resilience. The analysis highlights how chatbots can address key dimensions of the BANI framework and serve as adaptive instruments within pedagogical ecosystems undergoing transformation. The proposed architecture remains conceptual at this stage, providing a foundation for future empirical validation in real-world institutional settings.