Table 1
Overview of the methodological evolution, shared infrastructures, and main application areas of MIR, from music analysis and representation to interactive systems and generation.
Table 2
Indicative distinctions between exploratory research and commercial deployment in music‑AI; their boundaries vary across jurisdictions and use cases.
| Exploratory Research | Commercial Deployment | |
|---|---|---|
| Purpose | Scientific exploration, methodological development, and validation | Operational use, market delivery, and sustained user‑facing services |
| Scale and context | Prototypes, controlled studies, restricted access, and pre‑deployment testing; datasets and models may still be large | Systems placed on the market or put into service, often with broad access, continuous operation, and substantial impact |
| Legal context | May benefit from research exceptions or exclusions, including text and data mining provisions, fair‑use interpretations, and some pre‑deployment activities under the EU AI Act; other laws still apply | Subject to applicable licensing, contractual, consumer‑protection, risk‑management, and sector‑specific requirements |
| Responsibilities | Research integrity, documentation, reproducibility, ethical review, disclosure of limitations, and legal compliance | Licensing, accountability, user transparency, risk management, monitoring, consumer protection, and regulatory compliance |
| Evaluation | Scientific validity, robustness, generalization, and reproducibility | Real‑world reliability, user impact, safety, fairness, and ongoing monitoring |
| Failure risks | Invalid claims, misleading results, and wasted research resources | Harm to users, creators, communities, markets, and public trust |
Table 3
Proposed and interconnected principles for critically assessing current MIR research practices and guiding future work in trustworthy music‑AI.
Table 4
Six interconnected and mutually reinforcing priority areas for advancing trustworthy music‑AI that is open, culturally inclusive, and ethically grounded.
Figure 1
Conceptual framework connecting music‑AI challenges with principles for assessment, priorities for collective action, and the desired qualities of trustworthy music‑AI.