EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Counterfactual Audit of Ethnic-Origin Bias and Architectural Mitigation in Generative AI Assistants for Debt Collection in Savings and Credit Cooperatives

Edgar Hernán Changoluisa Lasluisa, María Belén Changoluisa Lasluisa, Pablo David Alemán Gualpa, Pablo M. Recalde Varela, Paul Francisco Baldeón-Egas

Computers · 01.11.2026 · 15 (11) · 765

DOI: 10.3390/computers15110765 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Computers, online erschienen am 01.11.2026. Verfasst von Edgar Hernán Changoluisa Lasluisa und 4 weiteren Autor:innen. Die Arbeit nennt 23 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Debt-collection automation requires assessment of differential treatment between equivalent profiles. This pre-deployment counterfactual audit evaluates tone severity and empathy in generative AI outputs for Ecuador’s Popular and Solidarity Financial Sector. Five models generated 2400 responses under unmitigated and fairness-instruction conditions (A/B), plus 400 surname-masked control tasks (C), across 40 synthetic scenarios and three surname pairs. None of the 12 primary contrasts remained statistically significant after Holm correction. This does not establish absence of bias, and the 40-scenario sample may limit detection of small effects. Results for instruction-based mitigation apply only to the tested Protocol V4 template; alternative formulations were not evaluated. Masking prevented surname exposure in all 400 control tasks without establishing general fairness or identical generated text. The primary human analysis used seven complete raters and 630 paired-item evaluations; three additional raters contributed 29 partial evaluations excluded from that analysis. Inter-rater agreement was near zero and judge–human alignment was weak, limiting interpretation of automated scores as measures of perceived severity or empathy. One complete rater was also a coauthor. The study provides scoped audit results and an exposure-control check rather than evidence of deployment effectiveness or general model fairness.

Zitieren

APA

Lasluisa, E. H. C., Lasluisa, M. B. C., Gualpa, P. D. A., Varela, P. M. R., & Baldeón-Egas, P. F. (2026). Counterfactual Audit of Ethnic-Origin Bias and Architectural Mitigation in Generative AI Assistants for Debt Collection in Savings and Credit Cooperatives. Computers, 15(11), 765. https://doi.org/10.3390/computers15110765

Vancouver

Lasluisa EHC, Lasluisa MBC, Gualpa PDA, Varela PMR, Baldeón-Egas PF. Counterfactual Audit of Ethnic-Origin Bias and Architectural Mitigation in Generative AI Assistants for Debt Collection in Savings and Credit Cooperatives. Computers. 2026;15(11):765. doi:10.3390/computers15110765

BibTeX
@article{Lasluisa2026,
  author = {Lasluisa, Edgar Hernán Changoluisa and Lasluisa, María Belén Changoluisa and Gualpa, Pablo David Alemán and Varela, Pablo M. Recalde and Baldeón-Egas, Paul Francisco},
  title = {Counterfactual Audit of Ethnic-Origin Bias and Architectural Mitigation in Generative AI Assistants for Debt Collection in Savings and Credit Cooperatives},
  journal = {Computers},
  year = {2026},
  volume = {15},
  number = {11},
  pages = {765},
  doi = {10.3390/computers15110765}
}
Kostenlos starten →