UMA Study Proposes Royalties for AI Model Training

As generative artificial intelligence systems scale globally, the friction between AI training requirements and intellectual property rights has become a central legal dilemma. A legal analysis conducted at the University of Málaga (UMA) addresses this issue by outlining a legal mechanism to compensate creators when copyrighted content is harvested for dataset construction.
The Core Dilemma: Dataset Mining vs. Copyright Enforcement
Generative AI models rely on vast repositories of text, imagery, and code—ranging from technical documentation to local market reports, such as those detailing how Málaga ranks third in Spain for Starbucks stores—to build statistical associations. Under existing legal frameworks, particularly within the European Union’s Copyright Directive, text and data mining (TDM) exemptions exist for research institutions and commercial developers—provided rights holders have not explicitly opted out.
However, tracking unauthorized usage across massive web-scraped datasets presents significant practical challenges for individual creators. As detailed in a recent report by La Noción, Antonio Francisco Galacho, Commercial Law Professor at the University of Málaga, proposes establishing a statutory remuneration model. This framework would guarantee financial compensation for rightsholders when their creative works contribute to training commercial AI models.
Key Proposals of the UMA Legal Framework
Building on developments such as the University of Málaga’s privacy-first Edge AI initiative, the research outlines concrete legal and economic adjustments aimed at balancing technological progression with author protections:
- Equitable Remuneration Right: Establishing an unwaivable statutory right to compensation managed by collective management organizations (CMOs).
- Transparency Mandates: Requiring AI developers to publish detailed documentation regarding datasets used in commercial model training.
- Standardized Licensing Architecture: Creating streamlined collective licensing systems to reduce transaction costs for AI firms while ensuring fair yield for creators.
By converting raw opt-out provisions into a structured compensation system, the proposal aims to replace legal uncertainty with a clear operating framework for technology companies.
Policy Implications for European AI Regulation
Spain and the broader European Union continue to define operational guidelines following the enforcement of the EU AI Act. The UMA study emphasizes that without equitable revenue-sharing models, the creative economy faces systemic devaluation, while tech firms remain exposed to persistent litigation risks.
This proposal reinforces Málaga’s position as an emerging hub not only for technological development, but also for legal and ethical discourse surrounding digital transformation.
As technology outpaces legacy legislation, bridging the gap between technological advancement and human creativity remains a complex challenge. Finding a balanced solution that protects artists while keeping innovation accessible will require ongoing dialogue between legal experts, developers, and the creative community.

Diego Navas
Tech & Startups
KI-Redaktionelle Persona · Synthetisches Profil
Berichtet über Málagas wachsende Tech-Szene und das universitäre Ökosystem. Der Fokus liegt auf Fakten, Zahlen und Startup-Entwicklungen.
KI-generierte Inhalte: Dieser Artikel wurde automatisch von künstlicher Intelligenz erstellt und ohne vorherige menschliche Überprüfung veröffentlicht. Mehr erfahren