@inproceedings{el-herraoui-etal-2026-half,
title = "{HALF}: Harm-Aware {LLM} Fairness Evaluation Aligned with Deployment",
author = "El Herraoui, Omar and
Mekky, Ali and
Nakov, Preslav and
Wang, Yuxia",
editor = "Zhang, Yue and
Xu, Ruifeng and
Purwarianti, Ayu and
Rong, Tong and
Suzuki, Jun",
booktitle = "Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the {A}sia-Pacific Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = nov,
year = "2026",
address = "Hengqin, China",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.aacl-long.22/",
pages = "427--452",
ISBN = "979-8-89176-439-2",
abstract = "Large language models (LLMs) are increasingly deployed across high-impact domains, from clinical decision support and legal analysis to hiring and education, making fairness and bias evaluation before deployment critical. However, existing evaluations lack grounding in real-world scenarios and do not account for differences in harm severity, e.g., a biased decision in surgery should not be weighed the same as a stylistic bias in text summarization. To address this gap, we introduce HALF (Harm-Aware LLM Fairness), a deployment-aligned framework that assesses model bias in realistic applications and weighs the outcomes by harm severity. HALF organizes evaluation datasets into harm tiers (Severe, Moderate, Mild) based on the specific task and bias type they measure. Harm severity is assigned according to deployment-relevant criteria grounded in regulatory and risk assessment frameworks. Our evaluation results across eight LLMs show that (1) LLMs are not consistently fair across datasets, (2) model size or performance do not guarantee fairness, and (3) reasoning models perform better in medical decision support but worse in education. We conclude that HALF exposes a clear gap between previous benchmarking success and model fairness. All data and code are available at https://github.com/mbzuai-nlp/HALF"
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="el-herraoui-etal-2026-half">
<titleInfo>
<title>HALF: Harm-Aware LLM Fairness Evaluation Aligned with Deployment</title>
</titleInfo>
<name type="personal">
<namePart type="given">Omar</namePart>
<namePart type="family">El Herraoui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ali</namePart>
<namePart type="family">Mekky</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Preslav</namePart>
<namePart type="family">Nakov</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yuxia</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ruifeng</namePart>
<namePart type="family">Xu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ayu</namePart>
<namePart type="family">Purwarianti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Tong</namePart>
<namePart type="family">Rong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Suzuki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Hengqin, China</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-439-2</identifier>
</relatedItem>
<abstract>Large language models (LLMs) are increasingly deployed across high-impact domains, from clinical decision support and legal analysis to hiring and education, making fairness and bias evaluation before deployment critical. However, existing evaluations lack grounding in real-world scenarios and do not account for differences in harm severity, e.g., a biased decision in surgery should not be weighed the same as a stylistic bias in text summarization. To address this gap, we introduce HALF (Harm-Aware LLM Fairness), a deployment-aligned framework that assesses model bias in realistic applications and weighs the outcomes by harm severity. HALF organizes evaluation datasets into harm tiers (Severe, Moderate, Mild) based on the specific task and bias type they measure. Harm severity is assigned according to deployment-relevant criteria grounded in regulatory and risk assessment frameworks. Our evaluation results across eight LLMs show that (1) LLMs are not consistently fair across datasets, (2) model size or performance do not guarantee fairness, and (3) reasoning models perform better in medical decision support but worse in education. We conclude that HALF exposes a clear gap between previous benchmarking success and model fairness. All data and code are available at https://github.com/mbzuai-nlp/HALF</abstract>
<identifier type="citekey">el-herraoui-etal-2026-half</identifier>
<location>
<url>https://aclanthology.org/2026.aacl-long.22/</url>
</location>
<part>
<date>2026-11</date>
<extent unit="page">
<start>427</start>
<end>452</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T HALF: Harm-Aware LLM Fairness Evaluation Aligned with Deployment
%A El Herraoui, Omar
%A Mekky, Ali
%A Nakov, Preslav
%A Wang, Yuxia
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-439-2
%F el-herraoui-etal-2026-half
%X Large language models (LLMs) are increasingly deployed across high-impact domains, from clinical decision support and legal analysis to hiring and education, making fairness and bias evaluation before deployment critical. However, existing evaluations lack grounding in real-world scenarios and do not account for differences in harm severity, e.g., a biased decision in surgery should not be weighed the same as a stylistic bias in text summarization. To address this gap, we introduce HALF (Harm-Aware LLM Fairness), a deployment-aligned framework that assesses model bias in realistic applications and weighs the outcomes by harm severity. HALF organizes evaluation datasets into harm tiers (Severe, Moderate, Mild) based on the specific task and bias type they measure. Harm severity is assigned according to deployment-relevant criteria grounded in regulatory and risk assessment frameworks. Our evaluation results across eight LLMs show that (1) LLMs are not consistently fair across datasets, (2) model size or performance do not guarantee fairness, and (3) reasoning models perform better in medical decision support but worse in education. We conclude that HALF exposes a clear gap between previous benchmarking success and model fairness. All data and code are available at https://github.com/mbzuai-nlp/HALF
%U https://aclanthology.org/2026.aacl-long.22/
%P 427-452
Markdown (Informal)
[HALF: Harm-Aware LLM Fairness Evaluation Aligned with Deployment](https://aclanthology.org/2026.aacl-long.22/) (El Herraoui et al., AACL-IJCNLP 2026)
ACL
- Omar El Herraoui, Ali Mekky, Preslav Nakov, and Yuxia Wang. 2026. HALF: Harm-Aware LLM Fairness Evaluation Aligned with Deployment. In Proceedings of the 15th International Joint Conference on Natural Language Processing and the 5th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 427–452, Hengqin, China. Association for Computational Linguistics.