Abstract:

The rapid adoption of generative artificial intelligence (AI) in higher education has largely been framed in terms of tool usage, prompt literacy, and productivity enhancement. However, such perspectives risk reducing AI integration to operational competence rather than epistemic transformation. This study proposes a Trilingual Framework for AI-enabled education, arguing that meaningful learning in socio-technical domains requires fluency across three interconnected epistemic “languages”: Design, AI, and Domain Knowledge.

The framework emerged from a qualitative study conducted in a graduate-level Smart City course where students used no-code generative AI platforms to design and prototype real-world applications within a three-week iterative cycle. Data were collected through longitudinal structured reflections at three time points and analyzed using an inductive–abductive coding approach informed by epistemic fluency and boundary crossing theory.

Findings indicate that while students rapidly developed operational fluency with AI tools, meaningful innovation required recursive translation between computational outputs, iterative design logic, and contextual disciplinary reasoning. Generative AI functioned as an epistemic mediator, accelerating prototyping while simultaneously exposing tensions between technical feasibility and domain nuance. Learning intensified at these boundary encounters, where students negotiated mismatches between AI-generated artifacts and real-world constraints.

The Trilingual Framework conceptualizes AI-enabled learning not as additive skill acquisition but as integrative epistemic negotiation. By positioning AI as a distinct epistemic domain alongside design practice and disciplinary expertise, the framework extends theories of epistemic fluency into the context of generative AI. The study contributes a transferable pedagogical model for embedding AI within discipline-based education and offers implications for preparing students to navigate complex socio-technical systems in AI-augmented futures.

Keywords:

Generative AI in Education, Epistemic Fluency, Design–AI Integration, Domain Knowledge.