Analysis of Emotional Characteristics in AI-Generated Music

Aug. 2026 – Present

Role: Researcher · Period: Aug. 2026 – Present · Programme: AP Research

This AP Research project analyses the emotional characteristics of AI-generated music, assessing how well AI music models across generations capture intended emotional effects.

  • Study design. Designing a study to analyse and assess how well AI music models across generations capture intended emotional effects.
  • Data collection. Collecting data on human-composed music, early-version AI music, and latest-version AI music.
  • Evaluation pipeline. Using a neutral, cross-validated pipeline to compare emotional capabilities across model versions.
  • Research focus. How successive model generations differ in their ability to express the emotion a piece is meant to convey, relative to human-composed music.