What question did this study set out to answer?

The research aims to create a comprehensive classification of algorithmic biases based on their causal origins.

March 7, 2026Open Access

Unified Critical Taxonomy of Algorithmic Biases: Structural Cartography, Differential Diagnosis, and Diachronic Dynamics

Key Points

The research aims to create a comprehensive classification of algorithmic biases based on their causal origins.
Classified algorithmic biases into three main categories: emergent, architecture, and implanted biases.
Developed an in-depth analysis of computational architecture bias with empirical evidence from technical literature.
Proposed a heuristic for differential diagnosis to identify dominant causal categories in specific cases.
Modeled the dynamic nature of algorithmic bias, illustrating feedback loops between categories.
Established a novel classification system for algorithmic biases that is both exhaustive and logically sound.
Demonstrated how optimization decisions in AI lead to biased outputs, viewed as neutral trade-offs.
Showed that algorithmic bias evolves dynamically, influenced by user interactions and feedback mechanisms.

Abstract

This article proposes and develops the Unified Critical Taxonomy of Algorithmic Biases (UCAB) as the most complete systematic contribution to the field of structural criticism of artificial intelligence, grounded in the research program of the Psychoanalysis of Technogenesis (PdT) (Beltrán, 2025). The UCAB makes four simultaneous original contributions that no preceding framework has achieved: (1) it classifies all subtypes of algorithmic bias by causal origin —not by symptom, effect, or affected group— into three universal and logically exhaustive categories: emergent bias, architecture bias, and implanted bias; (2) it develops with unprecedented depth the computational architecture bias as an autonomous subcategory, demonstrating with empirical evidence from technical literature (Frantar et al., 2022; Lin et al., 2024; Hooper et al., 2024; Liu et al., 2024) how optimization decisions —quantization, KV cache compression, context window limits, parameter pruning— produce epistemically impoverished responses that engineering treats as neutral trade-offs but that PdT reveals as second-order political acts; (3) it proposes a differential diagnosis heuristic that enables researchers and auditors to determine which causal category is dominant in a concrete case, operationalizing the taxonomy as a field instrument; and (4) it models the diachronic dynamics of algorithmic bias: the three feedback circuits between categories that demonstrate that algorithmic bias is not a static state but a dynamic system that self-replicates and amplifies over time through user interaction. The article is self-sufficient: it integrates PdT theoretical foundations, the complete state of the art, and an exhaustive catalog of 21 documented subtypes, requiring no prior reading of previous documents in the series

Unified Critical Taxonomy of Algorithmic Biases: Structural Cartography, Differential Diagnosis, and Diachronic Dynamics

Key Points

Abstract

Cite This Study