AI Governance &
Algorithmic
Accountability
Research
Isabel Velarde's research is published as practitioner working papers on AI governance, algorithmic accountability, and institutional oversight in Latin America, with a focus on high-informality environments where imported governance models routinely fail to address structural conditions specific to the region. Her current research program examines institutional asymmetries on both sides of evaluative AI deployment: who or what becomes admissible to evaluation before assessment occurs, and whether organizations retain sufficient capacity to govern the AI systems they operate once deployment occurs. The work proposes practitioner instruments calibrated for Latin American deployment conditions, including the «Pre-Evaluative Audit» («PEA») and the «Institutional AI Dependency Review» («IADR»).
Working Papers
Who the Algorithm Doesn't See
A Practitioner Audit for Pre-Evaluative Exclusion in High-Informality Economies (Latin America)
Working paper · June 2026 · Zenodo · CC BY 4.0 · DOI: 10.5281/zenodo.19665794
This practitioner working paper sets out the concepts of «Admissible Evidence Boundary» and «Pre-Evaluative Exclusion» to examine how algorithmic systems construct evaluability through the records, signals, and documentary forms they recognize as admissible evidence before any score, classification, or eligibility determination is produced. Focusing on Latin America and other high-informality economies, this paper analyzes how substantial portions of economically active populations may remain only partially visible within automated evaluative infrastructures despite participating continuously in markets, labor systems, and social life.
This paper proposes the «Pre-Evaluative Audit» («PEA»), a governance instrument structured around four institutional decision points: vendor disclosure, board oversight, investment due diligence, and public procurement. The framework is illustrated through five Latin American cases: SISBEN IV (Colombia), RappiCard (Mexico), Plataforma Tecnológica de Intervención Social (Argentina), Registro Social de Hogares (Chile), and SISFOH (Peru). This paper positions evaluability itself as an object of institutional oversight, extending AI governance upstream to the institutional conditions that determine who can be evaluated before algorithmic decision-making begins.
Read on Zenodo → doi.org/10.5281/zenodo.19665794
The Governance Gap Behind AI Deployment in High-Informality Economies
Explaining Institutional Dependency in
AI Adoption
Working paper · June 2026 · Zenodo · CC BY 4.0 · DOI: 10.5281/zenodo.20217978
This practitioner working paper examines the «Governance Deployment Gap», the divergence between an institution's ability to operate an AI system and its retained capacity to examine, challenge, reconstruct, and sustain that system independently over time. Successful deployment can conceal governance fragility: an organization may operate a system it cannot independently govern.
This paper names «Capability Self-Attribution», the mechanism by which an organization, relying on sustained access to externally supplied outputs, attributes to itself a capability it lacks, producing «Unrecognized Provider Dependence». It specifies «Capacity-Output Asymmetry» as the documentary signature of this condition, names «Procurement Governance Analysis» («PGA») as the method for identifying it in acquisition records, and sets out the «Institutional AI Dependency Review» («IADR»), a board-level instrument that surfaces this dependence at acquisition, before deployment. Six documented institutional cases, three high-formality and three high-informality, illustrate the framework as structural analogies rather than empirical proof.
Developed for high-informality economies in Latin America, the IADR complements the NIST AI RMF, ISO/IEC 42001, and the OECD AI Principles by focusing on retained governance capacity rather than system-level properties alone.
Read on Zenodo → doi.org/10.5281/zenodo.20217978
Open Access Research
The papers are published on Zenodo, the open-access repository operated by CERN, and released under Creative Commons Attribution 4.0 International (CC BY 4.0). This open-access approach ensures the research remains freely accessible to policymakers, investors, academics, and practitioners across the region, supporting open dialogue and evidence-based governance in contexts where commercial or institutional paywalls would otherwise limit reach. The research is indexed in OpenAIRE, citable through permanent DOIs, and linked to the author's ORCID Identifier (0009-0005-0798-0798).