Metraipolis › archive › dc_aa778f993d06

Benevolent Bias in Multi-Turn Human-Agent Dialogue — arXiv:2608.29206

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.29206
License (as archived)
CC0-1.0
Archived
2026-10-17T07:44:08.356Z
SHA-256 of the archived text
e3b41f07d63ccf61e2aefb3640688e2511eb0cfd263bd523bf73d0dba36f82ca
Stable URL of this page
https://metraipolis.com/documento/dc_aa778f993d06

How to cite

Benevolent Bias in Multi-Turn Human-Agent Dialogue — arXiv:2608.29206 — arXiv (en), https://arxiv.org/abs/2608.29206 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_aa778f993d06 (Metraipolis archive dc_aa778f993d06, sha256 e3b41f07d63ccf61e2aefb3640688e2511eb0cfd263bd523bf73d0dba36f82ca)
BibTeX
@misc{dcaa778f993d06,
  title = {Benevolent Bias in Multi-Turn Human-Agent Dialogue — arXiv:2608.29206},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.29206},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_aa778f993d06; sha256 e3b41f07d63ccf61e2aefb3640688e2511eb0cfd263bd523bf73d0dba36f82ca},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29206. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Bias in human-agent interaction can manifest not only through hostile language but also as benevolent bias, whereby unequal treatment hides behind a warm, positive tone. To make it detectable, we operationalise benevolent bias along two dimensions, tone and treatment, yielding three classes: neutral support, overt bias, and benevolent bias. Building on these definitions, we construct BENEVDIAL, a class-balanced corpus of 362,880 multi-turn support dialogues spanning user and agent demographics, roles, and generators, to support controlled evaluation. We then test two detector families on it: off-the-shelf safety detectors and prompted large language model (LLM) judges. Our findings reveal a notable detection gap: off-the-shelf detectors reliably flag overt bias yet largely fail to identify benevolent bias. LLM judges improve sensitivity when guided by explicit detection criteria, but this comes at the cost of increased misclassification of neutral supportive statements as benevolent bias, a tendency that is further exacerbated by the presence of demographic context. These findings suggest that fair monitoring of human-agent dialogue must look beyond surface cues to whether the agent's treatment is disparate.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.29206 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-17T07:28:26.524Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).