AI Resilience Report forArt, Drama, and Music Teachers, Postsecondary

Right now, AI in postsecondary arts education is mostly being used to augment teaching — not replace it. In a Music Educators Journal article published in February 2026, researcher Hyesoo Yoo argues that AI literacy is now essential for educators across all disciplines, framing tools like ChatGPT and music generators as new "instruments" professors must learn alongside their students. At Berklee College of Music, faculty and students are openly debating how far to go ^[1]: one composition senior describes professors in the film scoring department using generative AI to write musical cues ^[1], while more than 400 students have signed a petition opposing AI in their songwriting curriculum.

Oberlin's "Year of AI Exploration" ^[2] shows the augmentation pattern clearly — conservatory faculty ran workshops on computational creativity and music generation, and the college gave instructors licensed access to ChatGPT and Gemini rather than replacing courses. A 2025 systematic review of generative AI in art education ^[3] similarly found that tools like DALL-E and Midjourney are mostly being used as brainstorming and ideation aids, not as substitutes for instruction.

Adoption is uneven and slower than in other fields. EDUCAUSE's 2026 research on AI in higher-ed work ^[4] notes that institutions are still piloting AI rather than scaling it, and Inside Higher Ed's 2026 predictions ^[5] warn that a possible AI "bubble" could further slow campus rollouts. Arts faculty also face strong cultural pushback: Berklee professor Marti Epstein says students returned to campus less trustful of AI and are worried that careers in songwriting and film scoring will be overtaken by it.

Copyright concerns about training data, tenure-protected labor that isn't cheap to "replace," and the deeply human nature of mentorship, critique, and live performance all push adoption to be careful rather than aggressive. The good news for you: the irreplaceable parts of this job — coaching a nervous performer, leading a heated classroom discussion, curating a gallery show, showing up to a community concert — are exactly the tasks the data marks as least automatable, and they're the ones professors say matter most.