This study explores how advanced learners of Korean shape the language landscape by investigating the use of ‘-(u)lkka' in chat conversations. The study found that advanced learners of Korean use ‘-(u)lkka' for three different types of confirmation or suggestion on mobile platforms: 1) yes/no confirmation of previously shared information; 2) wh-confirmation, confirming new information and the interlocutor's intentions; and 3) performative suggestions, proposing ideas for buying or selling items, confirming the interlocutor's intentions on the surface level, while implying the speakers' performativity on a deeper level. Interestingly, when multiple instances of ‘-(u)lkka' occur in a single chat conversation, different confirmation strategies are employed, indicating contextual adaptation rather than a fixed approach. These findings reveal the dynamic transformation of the Korean language landscape, focusing on its adaptation to digital platforms and the evolving roles of language agents. Notably, conventional linguistic expressions are shifting towards mobile platforms, expanding the spectrum of language agents to include both language learners and native speakers. This shift empowers learners, enabling them to assume an active role as language agents, no longer dependent solely on native speakers. The study also recognizes the growing significance of artificial intelligence (AI) in contemporary language communication. The findings can contribute to enhancing AI systems and fostering more sophisticated human–AI interactions.