International taxation and AI
Identification of Unconventional Patterns in Double Taxation Agreements: An Analysis Based on AI-Assisted Natural Language Processing Models
Business Administration dissertation · Complutense University of Madrid · Distinction with honours (Top 1%) · UCM University Entrepreneurship Award 2025
A comparative analysis of the double taxation agreements concluded by Spain with a number of economically relevant countries, tested against the 2017 OECD Model Tax Convention, in order to identify atypical clauses and unconventional patterns arising from specific bilateral negotiations.
The study applies natural language processing techniques and prompt engineering over a large language model, assessing its capacity to systematise comparative analysis and to detect divergences from the model text.
The work was awarded a distinction with honours (Top 1%) and was recognised at the 13th UCM University Entrepreneurship Awards 2025 for its potential as a legal technology project.
Sources
- OECD Model Tax Convention on Income and on Capital (2017).
- Double taxation agreements concluded by Spain.
How to cite this work
DÍAZ-HEREDERO LÓPEZ, G., “Identification of Unconventional Patterns in Double Taxation Agreements: An Analysis Based on AI-Assisted Natural Language Processing Models”, DÍAZ-HEREDERO Legal & Tax, 2025. Available at: https://diazheredero-web.pages.dev/en/research/unconventional-patterns-double-taxation-agreements/ (accessed: dd/mm/aaaa).
Published content is informative in nature and does not constitute individual advice. It does not replace the analysis of a specific case or the review of the corresponding documentation.