Colusión algorítmica y derecho de la competencia: hacia un modelo regulatorio tecnológicamente adaptativo para los mercados digitales

The emergence of artificial intelligence in contemporary markets has generated a structural transformation in competitive logic, shifting the traditional axis of antitrust law from human volition toward automated technical agency. This research critically analyzes the challenges that algorithmic collusion poses to competition law, examining the insufficiency of classical legal categories (agreement, intent, culpability) when confronted with machine learning systems capable of converging on anticompetitive strategies without explicit communication. Through comparative analysis of European and American regulatory responses, three structural gaps are identified: normative, procedural, and institutional. The research proposes an adaptive regulatory model grounded in three pillars: technological attribution based on reasonable foreseeability of harm, algorithmic traceability through scalable technical audits, and institutional proportionality that adjusts regulatory requirements to each jurisdiction's capacity. The findings demonstrate that algorithmic collusion constitutes a historical turning point requiring abandonment of the anthropocentric paradigm of responsibility and adoption of technical fault criteria in eligendo and in vigilando. Regulatory effectiveness depends on constructing a legal architecture that articulates substantive, procedural, and organizational reform, ensuring coherence between normative ambition and technical viability in contexts of institutional asymmetry.