Emerging consensus suggests faculty should teach students to use large language models (LLMs) rather than ban them, but it is not clear that students need detailed AI-related instruction. To investigate, we conducted two studies: Study 1 used survey and focus group methods to assess how such instruction influenced students’ perceptions, while Study 2 used rater evaluation to examine how AI use affected message quality. Study 1 found no meaningful impact on perceptions. Study 2 found that instruction did not affect ratings, but genAI use did—messages composed with LLM assistance received higher evaluations than those without it. We conclude with recommendations for genAI-focused classroom instruction.