<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="choi-etal-2025-unintended">
    <titleInfo>
        <title>Unintended Harms of Value-Aligned LLMs: Psychological and Empirical Insights</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Sooyung</namePart>
        <namePart type="family">Choi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jaehyeok</namePart>
        <namePart type="family">Lee</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xiaoyuan</namePart>
        <namePart type="family">Yi</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Jing</namePart>
        <namePart type="family">Yao</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Xing</namePart>
        <namePart type="family">Xie</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">JinYeong</namePart>
        <namePart type="family">Bak</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2025-07</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Wanxiang</namePart>
            <namePart type="family">Che</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Joyce</namePart>
            <namePart type="family">Nabende</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ekaterina</namePart>
            <namePart type="family">Shutova</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Mohammad</namePart>
            <namePart type="given">Taher</namePart>
            <namePart type="family">Pilehvar</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Vienna, Austria</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-251-0</identifier>
    </relatedItem>
    <abstract>The application scope of Large Language Models (LLMs) continues to expand, leading to increasing interest in personalized LLMs that align with human values. However, aligning these models with individual values raises significant safety concerns, as certain values may correlate with harmful information. In this paper, we identify specific safety risks associated with value-aligned LLMs and investigate the psychological principles behind these challenges. Our findings reveal two key insights. (1) Value-aligned LLMs are more prone to harmful behavior compared to non-fine-tuned models and exhibit slightly higher risks in traditional safety evaluations than other fine-tuned models. (2) These safety issues arise because value-aligned LLMs genuinely generate text according to the aligned values, which can amplify harmful outcomes. Using a dataset with detailed safety categories, we find significant correlations between value alignment and safety risks, supported by psychological hypotheses. This study offers insights into the “black box” of value alignment and proposes in-context alignment methods to enhance the safety of value-aligned LLMs.</abstract>
    <identifier type="citekey">choi-etal-2025-unintended</identifier>
    <identifier type="doi">10.18653/v1/2025.acl-long.1532</identifier>
    <location>
        <url>https://aclanthology.org/2025.acl-long.1532/</url>
    </location>
    <part>
        <date>2025-07</date>
        <extent unit="page">
            <start>31742</start>
            <end>31768</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Unintended Harms of Value-Aligned LLMs: Psychological and Empirical Insights
%A Choi, Sooyung
%A Lee, Jaehyeok
%A Yi, Xiaoyuan
%A Yao, Jing
%A Xie, Xing
%A Bak, JinYeong
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-251-0
%F choi-etal-2025-unintended
%X The application scope of Large Language Models (LLMs) continues to expand, leading to increasing interest in personalized LLMs that align with human values. However, aligning these models with individual values raises significant safety concerns, as certain values may correlate with harmful information. In this paper, we identify specific safety risks associated with value-aligned LLMs and investigate the psychological principles behind these challenges. Our findings reveal two key insights. (1) Value-aligned LLMs are more prone to harmful behavior compared to non-fine-tuned models and exhibit slightly higher risks in traditional safety evaluations than other fine-tuned models. (2) These safety issues arise because value-aligned LLMs genuinely generate text according to the aligned values, which can amplify harmful outcomes. Using a dataset with detailed safety categories, we find significant correlations between value alignment and safety risks, supported by psychological hypotheses. This study offers insights into the “black box” of value alignment and proposes in-context alignment methods to enhance the safety of value-aligned LLMs.
%R 10.18653/v1/2025.acl-long.1532
%U https://aclanthology.org/2025.acl-long.1532/
%U https://doi.org/10.18653/v1/2025.acl-long.1532
%P 31742-31768

Markdown (Informal)

[Unintended Harms of Value-Aligned LLMs: Psychological and Empirical Insights](https://aclanthology.org/2025.acl-long.1532/) (Choi et al., ACL 2025)

• Unintended Harms of Value-Aligned LLMs: Psychological and Empirical Insights (Choi et al., ACL 2025)

ACL

• Sooyung Choi, Jaehyeok Lee, Xiaoyuan Yi, Jing Yao, Xing Xie, and JinYeong Bak. 2025. Unintended Harms of Value-Aligned LLMs: Psychological and Empirical Insights. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 31742–31768, Vienna, Austria. Association for Computational Linguistics.