Meta-identity and digital platforms: disputes over transparency and control in the algorithmic creation of identities by AI systems

Abstract

This study applies the concept of meta-identity as an analytical key to understanding classificatory data processes, power dynamics, and governance structures in the contemporary digital ecosystem. Drawing on the analysis of three cases that generated significant social upheaval in Brazil, it investigates how identities on digital platforms are constructed, assigned, and operationalized by algorithmic systems, revealing tensions between subjectivity, technique, and normativity. Based on both classical and contemporary approaches to identity, the text seeks to articulate a theoretical proposal still in development: a meta-identity model that allows for mapping the internal flows of digital classificatory systems — their input vectors, procedural dynamics, and outputs. Acknowledging the limits and challenges of any modeling attempt, this proposal is presented with scientific humility, as an initial effort to systematize complex and constantly evolving phenomena. The text also addresses the legal-normative governance that legitimizes such processes, proposing a critical methodology for empirical analysis and auditing of classificatory systems. Finally, it presents a systemic interpretation of the meta-identity model, highlighting its impacts on the definition of subjects, the production of truth, and the distribution of power in the digital sphere.

Portuguese Version: https://doi.org/10.1590/SciELOPreprints.13161