exploring the (non)-impact of artificial intelligence robots on learning medical english

Recent developments in artificial intelligence (ai) robots have led to a range of educational technologies designed to support standard language varieties. however, the ability of these robots’ knowledge bases to create appropriate contexts for university language learning is still not well understood. this quasi-experimental study, conducted during the 2024 academic year, involved 1205 students from different academic majors from different first languages. using a between-students design, we evaluated the effectiveness of learning english through ai robots with and without custom-built knowledge bases, employing mini clinical evaluation exercise for assessment. the analysis of the gathered data through generalized estimating equations suggested that students using custom-built ai robots became more socially accountable in medical english than those using robots with non-custom-built knowledge bases. this improvement was especially notable among students in fields requiring direct interaction with audiences, who reported a more positive learning experience. based on these findings, we provide recommendations for researchers and educators on effectively incorporating ai robots into university language education, emphasizing the potential for improved learning experiences through custom-built technologies.