Artificial Intelligence (AI) is revolutionizing how scientists and professional agricultural and natural resources (ANR) communicators and educators engage audiences. Land grant universities have embraced podcasts as a movement to connect listeners to ANR content. AI tools like Descript have infiltrated podcasting for text-based audio production to quickly add, remove, and replace words with machine-generated avatar voices. ANR communication courses must prepare upcoming science communication professionals to explore AI. We will present the design, implementation, and findings of a one-year case study across two classes at the University of Florida (UF) and one class at Kansas State University (KSU) to examine ANR communication students’ perceptions of AI, self-efficacy using AI podcasting tools, and trust in AI-human generated podcasts. Data collection included a) pre/post-student interviews, b) pre/post-student surveys, c) post-podcast guest surveys, and d) final student-AI-produced podcasts. Qualitative data were open, axial, and thematically coded. Quantitative data were analyzed for descriptive statistics. The pre/post-student surveys included Likert scale questions averaged into four constructs of AI perceptions, podcasting self-efficacy, AI self-efficacy, and AI trust. Results from the spring 2024 UF pilot (n = 16) indicated the projected-based learning podcasting course positively influenced students’ scores in all categories. However, the lowest increase occurred in AI trust (pre-M 3.55 pre-SD 0.43/ post-M 3.57 / post-SD 0.45). Results from the fall 2024 KSU launch (n = 23) also showed the course increased student scores in all categories with the greatest growth in AI self-efficacy (pre-M 2.46 pre-SD 1.08/ post-M 3.88 / post-SD 0.73). Qualitative results revealed students foresee the future AI uses to include ideation and task efficiency in ANR communication careers, as well as writing and graphic design. This project was funded by the Association of Public Land Grant Universities Innovative Teaching Award (2024).