@inproceedings{sauter-schirmer-2026-context,
title = "On the Context Sensitivity of {LLM} Moral Judgment",
author = "Sauter, Adrian and
Schirmer, Mona",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.178/",
pages = "4068--4113",
ISBN = "979-8-89176-459-0",
abstract = "A human{'}s moral decision depends heavily on the context. Yet research on LLM morality has largely studied fixed scenarios. We address this gap by introducing Contextual MoralChoice, a dataset of moral dilemmas with systematic contextual variations known from moral psychology to shift human judgment: consequentialist, emotional, and relational. Evaluating 22 LLMs, we find that nearly all models are context-sensitive, shifting their judgments towards rule-violating behavior. Comparing with a human survey, we find that models and humans are most triggered by different contextual variations, and that a model aligned with human judgments in the base case is not necessarily aligned in its contextual sensitivity. This raises the question of controlling contextual sensitivity, which we address with an activation steering approach that can reliably increase or decrease a model{'}s contextual sensitivity. Code and data: \url{https://github.com/adrian-sauter/contextual_moralchoice}."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sauter-schirmer-2026-context">
<titleInfo>
<title>On the Context Sensitivity of LLM Moral Judgment</title>
</titleInfo>
<name type="personal">
<namePart type="given">Adrian</namePart>
<namePart type="family">Sauter</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mona</namePart>
<namePart type="family">Schirmer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>A human’s moral decision depends heavily on the context. Yet research on LLM morality has largely studied fixed scenarios. We address this gap by introducing Contextual MoralChoice, a dataset of moral dilemmas with systematic contextual variations known from moral psychology to shift human judgment: consequentialist, emotional, and relational. Evaluating 22 LLMs, we find that nearly all models are context-sensitive, shifting their judgments towards rule-violating behavior. Comparing with a human survey, we find that models and humans are most triggered by different contextual variations, and that a model aligned with human judgments in the base case is not necessarily aligned in its contextual sensitivity. This raises the question of controlling contextual sensitivity, which we address with an activation steering approach that can reliably increase or decrease a model’s contextual sensitivity. Code and data: https://github.com/adrian-sauter/contextual_moralchoice.</abstract>
<identifier type="citekey">sauter-schirmer-2026-context</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.178/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>4068</start>
<end>4113</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T On the Context Sensitivity of LLM Moral Judgment
%A Sauter, Adrian
%A Schirmer, Mona
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F sauter-schirmer-2026-context
%X A human’s moral decision depends heavily on the context. Yet research on LLM morality has largely studied fixed scenarios. We address this gap by introducing Contextual MoralChoice, a dataset of moral dilemmas with systematic contextual variations known from moral psychology to shift human judgment: consequentialist, emotional, and relational. Evaluating 22 LLMs, we find that nearly all models are context-sensitive, shifting their judgments towards rule-violating behavior. Comparing with a human survey, we find that models and humans are most triggered by different contextual variations, and that a model aligned with human judgments in the base case is not necessarily aligned in its contextual sensitivity. This raises the question of controlling contextual sensitivity, which we address with an activation steering approach that can reliably increase or decrease a model’s contextual sensitivity. Code and data: https://github.com/adrian-sauter/contextual_moralchoice.
%U https://aclanthology.org/2026.emnlp-main.178/
%P 4068-4113
Markdown (Informal)
[On the Context Sensitivity of LLM Moral Judgment](https://aclanthology.org/2026.emnlp-main.178/) (Sauter & Schirmer, EMNLP 2026)
ACL
- Adrian Sauter and Mona Schirmer. 2026. On the Context Sensitivity of LLM Moral Judgment. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 4068–4113, Budapest, Hungary. Association for Computational Linguistics.