/ News Posts / Five Assignments That Keep the Learning Human
Remembering the Music Standards Anchors That AI Cannot Touch
By NAfME Member Dr. José Valentino Ruiz
Every subject has an AI article now. Most of them say the same thing: Set a policy, teach students to cite the tool, stay vigilant. Music needs a different article, because music has something other subjects do not: You can hear whether learning happened. A polished essay can hide an absent mind. A performance cannot. When a student plays, the room hears exactly what the student knows, and what the student skipped. That makes your classroom the best place in the school to answer the real question about AI, which is not whether students will use it, but whether they will still learn when they do.
The Newest Evidence, and What It Actually Says
The OECD’s Digital Education Outlook 2026, released in January 2026, synthesized the international research on generative AI in classrooms, and its central finding is blunt: AI can dramatically improve student performance while leaving learning behind. The clearest evidence comes from a field experiment in Türkiye with high school students. Access to GPT-4 improved practice performance by about 48 percent with a standard chatbot, and by about 127 percent with a tutoring version designed to support learning. Then researchers took the tool away and gave a closed-book exam. Students who had used the standard chatbot scored about 17 percent worse than students who had never used AI at all. Only the tutoring version preserved the gains (Bastani et al., 2024; OECD, 2026). The practice set looked great. The understanding never formed.
The same pattern shows up in other recent research. At the MIT Media Lab, researchers asked students to write essays, some with an AI assistant and some without. Afterward, 83 percent of the students who used AI could not quote from their own essay (Kosmyna et al., 2025). A 2025 study found heavier reliance on AI tools was associated with lower critical thinking scores (Gerlich, 2025). When a tool does the cognitive work, the mind does less of it, and the skill quietly atrophies.
The music-specific news is newer, and more hopeful, with a condition attached. In a 2026 study published in Frontiers in Psychology, Zhuang and Li studied 405 university students in China doing collaborative music creation with generative AI, and they found the AI-supported collaboration significantly raised creative interest, creative self-efficacy, self-regulated learning, and perceived creative competence. Read the condition carefully: The students were working together, and the AI was supporting their creation, not replacing it. The gains came from the collaboration, not from the tool alone. That distinction is the whole article.
UNESCO’s 2024 competency frameworks for students and teachers point the same way. Launched in September 2024, the frameworks put a human-centered mindset first: Students should understand and assert their agency in relation to AI, moving through Understand, Apply, and Create, while teachers progress through Acquire, Deepen, and Create (Miao & Cukurova, 2024; Miao et al., 2024). The frameworks treat students as co-creators, not consumers. Your classroom is where that idea gets tested, one assignment at a time.
Why Music Is the Exception That Proves the Rule
In most subjects, the gap between performance and learning is invisible until the exam. In music, it is audible in the first eight bars. A student who lets a tool write the melody cannot sing it back. A student who lets a tool choose the chords cannot explain why the chords go there. The product exists. The learning does not.
That is why the ear is the curriculum. Lucy Green (2008) studied informal music learning across 21 schools and more than 1,500 students and found that young musicians learn best when they choose the music, learn mainly by ear, learn with peers, and keep listening, performing, improvising, and composing connected instead of splitting them apart. Every one of those conditions is threatened when AI steps into the middle of the process, because the tool separates the student from the sound. Your job is to keep the student close to the sound.
The research on music and young people supports the same priority. Active engagement with music affects concentration, self-confidence, emotional sensitivity, and social skills, and those benefits appear when the experience is enjoyable and rewarding (Hallam, 2010). A classroom where students listen hard, make choices, and hear the results is a classroom where those benefits show up. A classroom where a tool makes the choices is one where they do not.
The National Core Arts Standards give you the frame. Music learning rests on four anchors: Creating, Performing, Responding, and Connecting (National Coalition for Core Arts Standards, 2014). Only the first two involve making sound. Responding and Connecting are about listening, interpreting, and placing music in a human context. Those are the anchors AI cannot touch, and they are the ones most worth protecting. A student who can respond to music, who can say what a piece does and why it matters, has something no tool can generate for them.
Five Assignments That Keep the Learning Human
Here is the part that matters most. Not a policy. Assignments. Each one is built so the learning stays with the student even when a tool is in the room.
First, the sing-it-back rule. Any melody, harmony, or lyric a student brings in from an AI tool must be sung or played back by the student, from memory, before it enters a piece. If the student cannot reproduce it, the student does not own it yet, and the work is not ready. The rule is simple, it costs nothing, and it makes the Türkiye finding visible in your room: The student who can sing it back learned it, and the student who cannot did not, no matter how good the product sounds.
Second, the AI-free first draft. When students compose, the first four bars come from the ear alone: no tools, no suggestions, just the student and the sound. After that, AI may enter as a source of variations, not as the author. Students compare the tool’s ideas with their own and defend their choices out loud. The comparison is the lesson. A student who can say why their own four bars are better than the tool’s twelve has learned judgment, which is the skill the tool cannot supply.
Third, the listening log. Once a week, each student writes three sentences about something heard in rehearsal: a balance problem, a phrase that worked, a moment when the ensemble locked in. The log trains attention, and attention is the muscle that heavy AI use weakens. Collect the logs. They are the most honest assessment data you will gather all term, because they show you what each student actually hears.
Fourth, the tutor version. This one comes directly from the Türkiye experiment. The tutoring version of the tool preserved learning because it asked questions instead of handing over answers. Set that as the rule for AI use in your room: The tool may question, never answer. It can ask what the student hears in a phrase, suggest what to listen for, or quiz the student on a chord choice. It may not write the phrase, choose the chord, or finish the thought. Students learn quickly that a tool that only asks questions is still useful, and they learn something better: the difference between help that teaches and help that replaces.
Fifth, the class policy, written with your students, not for them. Three columns: AI welcome, AI with limits, AI off limits. Post it on the wall. Revisit it each quarter. The conversation about the policy teaches more about judgment than any single assignment can, and it gives students the agency UNESCO’s frameworks call for: They are deciding, together, where the tool belongs.
The Work Ahead
AI will keep getting better at the easy parts of music: generating material, answering questions, and producing polished surfaces. Your job is to protect the hard parts: listening, judging, choosing, and making sound with other people in a room. Those parts build a musician, and they build a person who can think. If your students leave able to hear deeply, choose deliberately, and create something unmistakably theirs, then AI has been a tool in your room, not a replacement for it. That is the standard. Hold it.
References
- Bastani, H., et al. (2024). Generative AI without guardrails can harm learning. Field experiment in Türkiye, as synthesized in OECD (2026).
- Gerlich, M. (2025). AI tools in society: Impacts on cognitive offloading and the future of critical thinking. Societies.
- Green, L. (2008). Music, informal learning and the school: A new classroom pedagogy. Ashgate.
- Hallam, S. (2010). The power of music: Its impact on the intellectual, social and personal development of children and young people. International Journal of Music Education, 28(3), 269-289.
- Kosmyna, N., et al. (2025). Your brain on ChatGPT: Accumulation of cognitive debt when using an AI assistant for essay writing task. MIT Media Lab.
- Miao, F., & Cukurova, M. (2024). AI competency framework for teachers. UNESCO.
- Miao, F., Shiohira, K., & Lao, N. (2024). AI competency framework for students. UNESCO.
- National Coalition for Core Arts Standards. (2014). National Core Arts Standards.
- OECD. (2026). OECD Digital Education Outlook 2026: Exploring Effective Uses of Generative AI in Education. OECD Publishing.
- Zhuang, Z., & Li, X. (2026). The influence of collaborative music creation supported by generative artificial intelligence on students’ creativity. Frontiers in Psychology, 16, 1709513.
About the Author
NAfME member Dr. José Valentino Ruiz is a multi-Latin GRAMMY®-winning and multi-Global Music® Award-winning bassist, multi-instrumentalist, composer, producer, and educator. He is an Associate Professor and Founder/Coordinator of Music Business & Entrepreneurship at the University of Florida, a U.S. Department of State U.S. Speaker on Creative Economy and Global Arts Entrepreneurship, and an internationally active recording and performing artist. Ruiz also serves in editorial, research, and interdisciplinary leadership roles spanning music, creative industries, entrepreneurship, artificial intelligence, healthcare, and cultural diplomacy.
Did this blog spur new ideas for your music program? Share them on Connect! Interested in reprinting this article? Please review the reprint guidelines.
The National Association for Music Education (NAfME) provides a number of forums for the sharing of information and opinion, including blogs and postings on our website, articles and columns in our magazines and journals, and postings to our Connect member portal. Unless specifically noted, the views expressed in these media do not necessarily represent the policy or views of the Association, its officers, or its employees.
Published Date
November 19, 2026
Category
- Innovation
- Social Emotional Learning
- Technology
Copyright
November 19, 2026. © National Association for Music Education (NAfME.org)






