The Technological Infrastructure of the State Tax Administration Agency: Mass Data Processing Systems and Artificial Intelligence Techniques

Abstract

The Spanish State Tax Administration Agency (AEAT) has developed a sophisticated ecosystem of artificial intelligence (AI) and big data, whose actual scope and operations are marked by considerable opacity. This article overcomes this barrier through an empirical analysis of 1,731 public procurement files (from the 2015-2025 period), of which 77 are highly relevant, allowing for the reconstruction of the Agency’s technological architecture.

The research identifies and systematizes the key tools for mass data processing, predictive risk profiling, fraud network analysis, open-source intelligence (OSINT), and forensic analysis.

A significant gap is revealed between the AEAT’s official discourse, which focuses on assistance AI, and the contractual reality, which demonstrates massive and sustained investment in powerful control and surveillance systems.

This study provides the first detailed, empirically grounded mapping of this ecosystem, offering an indispensable factual basis for future legal analysis of the adequacy of safeguards and taxpayer rights in the face of these new technological capabilities.