Reframing L2 Pragmatic Competence in the Generative AI Era: A Co-Design-Oriented Critical Curriculum

The rapid incorporation of generative artificial intelligence (GAI) into second-language (L2) education has reopened a foundational question in language pedagogy: what should count as successful communication when English is routinely used among speakers who do not share a native-speaker norm? Much L2 pragmatic instruction continues to treat pragmatics as the acquisition of socially appropriate forms modelled on idealised native-speaker practices. This conceptual review article proposes a co-design-oriented critical pragmatic curriculum that positions GAI (implemented primarily through large language models, or LLMs) not as an infallible tutor but as a contestable intercultural interlocutor. Drawing on English as a Lingua Franca (ELF), World Englishes, translanguaging, critical pragmatics, critical digital literacy, and design-based research (DBR), the paper advances a unified pedagogical architecture comprising three modules: Persona Engineering and Dialectic Diversity, Real-Time Meaning Negotiation, and Deconstructing Algorithmic Bias, each enacted through a four-stage learning episode (persona engineering → GAI interaction → critical negotiation → reflection and transfer) and organised within an iterative co-design cycle. The paper contributes a theoretically grounded model for integrating GAI with intercultural and critical language pedagogy while retaining teacher mediation, learner agency, multilingual resources, and ethical safeguards. It argues that the educational value of GAI lies less in making learners sound more native-like than in helping them become more observant, adaptive, critical, and strategically competent participants in multilingual communication.