- Pat Pataranutaporn, Massachusetts Institute of Technology, Cambridge, MA, USA
- Nattavudh Powdthavee, Nanyang Technological University, Singapore, and IZA, University of Bonn, Germany
- Pattie Maes, Massachusetts Institute of Technology, Cambridge, MA, USA
Abstract
Surnames often convey implicit markers of social status, wealth, and lineage, shaping perceptions in ways that can perpetuate systemic biases and intergenerational inequality. This study is the first of its kind to investigate whether and how surnames influence AI-driven decision-making, focusing on their effects across key areas such as hiring recommendations, leadership appointments, and loan approvals. Using 72,000 evaluations of 600 surnames from the United States and Thailand—two countries with distinct sociohistorical contexts and surname conventions—we classify names into four categories: “Rich,” “Legacy,” “Normal,” and phonetically similar “Variant” groups. Our findings show that elite surnames consistently increase AI-generated perceptions of power, intelligence, and wealth, which in turn influence AI-driven decisions in high-stakes contexts. Mediation analysis reveals perceived intelligence as a key mechanism through which surname biases influence AI decision-making process. While providing objective qualifications alongside surnames mitigates most of these biases, it does not eliminate them entirely—especially in contexts where candidate credentials are low. These findings highlight the need for fairness-aware algorithms and robust policy measures to prevent AI systems from reinforcing systemic inequalities tied to surnames—an often-overlooked bias compared to more salient characteristics such as race and gender. Our work calls for a critical reassessment of algorithmic accountability and its broader societal impact, particularly in systems designed to uphold meritocratic principles while counteracting the perpetuation of intergenerational privilege.
Research Methodology
- The study analyzed 72,000 evaluations across 600 surnames from the United States and Thailand.
- AI-driven assessments were conducted using GPT-4o-mini, measuring perceived intelligence, wealth, power, commonality, and AI-driven decision-making, focusing on their effects across key areas such as hiring recommendations, leadership appointments, and loan approvals
- Multiple statistical methods, including OLS regression, mediation analysis, and bootstrap standard errors, ensured robust findings.
Key Findings
Surname Influence on AI Perceptions & Real World Impacts
- Elite surnames (Rich & Legacy) significantly increase AI ratings of intelligence, wealth, and power, which in turn influence AI-driven decisions in high-stakes contexts.
- In Thailand, elite surnames had a stronger impact on AI perception, likely due to unique surname laws ensuring surname exclusivity.
- AI was more likely to recommend individuals with elite surnames for executive roles, leadership, and high-profile positions.
- Entry-level hiring: Elite surnames had a negative effect, likely because AI associated these surnames with high status and deemed them “overqualified.”
- Loan approvals and political career recommendations were also influenced by surname-based perceptions.
- Variant surnames (phonetically similar but not identical) also triggered bias, particularly in intelligence assessments.
Intelligence as the Key Mediator
- Perceived intelligence was the primary driver of AI biases, linking surname prestige to decision-making outcomes.
- Mediation analysis confirmed that surname effects operate independently of actual qualifications.
Mitigation Through Objective Qualifications
- Adding academic and professional credentials reduced, but did not eliminate, surname bias.
- When candidate credentials were weak, AI defaulted to surname-based assumptions, reinforcing status-driven inequality.