Intersectional Sycophancy: How Perceived User Demographics Shape False Validation in Large Language Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915975304052736 |
|---|---|
| author | Maltbie, Benjamin Raval, Shivam |
| author_facet | Maltbie, Benjamin Raval, Shivam |
| contents | Large language models exhibit sycophantic tendencies, but whether this behavior varies systematically with perceived user demographics is underexplored. Inspired by intersectionality (overlapping identities produce compounded effects), we probe whether frontier models conditionally exhibit sycophancy. Across 768 multi-turn conversations spanning 128 personas (varying race, age, gender, confidence) and three domains (mathematics, philosophy, conspiracy theories), we find that sycophancy varies sharply with target model and domain, and emerges from combinations of perceived user traits rather than any single dimension. GPT-5-nano scores far higher than Claude Haiku 4.5 (average sycophancy scores of $\bar{x}=2.96$ vs.\ $1.74$, $p < 10^{-32}$); within GPT-5-nano, philosophy elicits 41\% more sycophancy than mathematics and Hispanic personas receive the highest scores across races. The worst-scoring persona, a confident, 23-year-old Hispanic woman, averages 5.33/10 (max 6/10), while Claude Haiku 4.5 remains uniformly low with no significant demographic variation. We argue that safety evaluations should incorporate identity-aware adversarial testing. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_11609 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Intersectional Sycophancy: How Perceived User Demographics Shape False Validation in Large Language Models Maltbie, Benjamin Raval, Shivam Artificial Intelligence Human-Computer Interaction Large language models exhibit sycophantic tendencies, but whether this behavior varies systematically with perceived user demographics is underexplored. Inspired by intersectionality (overlapping identities produce compounded effects), we probe whether frontier models conditionally exhibit sycophancy. Across 768 multi-turn conversations spanning 128 personas (varying race, age, gender, confidence) and three domains (mathematics, philosophy, conspiracy theories), we find that sycophancy varies sharply with target model and domain, and emerges from combinations of perceived user traits rather than any single dimension. GPT-5-nano scores far higher than Claude Haiku 4.5 (average sycophancy scores of $\bar{x}=2.96$ vs.\ $1.74$, $p < 10^{-32}$); within GPT-5-nano, philosophy elicits 41\% more sycophancy than mathematics and Hispanic personas receive the highest scores across races. The worst-scoring persona, a confident, 23-year-old Hispanic woman, averages 5.33/10 (max 6/10), while Claude Haiku 4.5 remains uniformly low with no significant demographic variation. We argue that safety evaluations should incorporate identity-aware adversarial testing. |
| title | Intersectional Sycophancy: How Perceived User Demographics Shape False Validation in Large Language Models |
| topic | Artificial Intelligence Human-Computer Interaction |
| url | https://arxiv.org/abs/2604.11609 |