@inproceedings{kang-etal-2026-q,
title = "{Q}-{CAT}: A Non-Parametric Query Filtering System for Prohibited Content in Dynamic Enterprise Environments",
author = "Kang, Juheon and
Choi, Eunkyung and
Oh, Hongseok and
Shin, Hyunseo and
Son, JuYoun and
Bae, Yunjin and
Hwang, Wonseok",
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.1221/",
pages = "27870--27891",
ISBN = "979-8-89176-459-0",
abstract = "Large Language Models (LLMs) are increasingly being adopted in enterprise environments, where employees may intentionally or accidentally submit prompts containing information that could cause legal or economic harm if leaked. Organizations may define categories of information (policies) that are prohibited from being submitted to commercial LLMs, but these definitions can vary between companies and evolve over time. This variability makes parametric classifiers less practical, as they require retraining whenever policies change. Here, we present Q-CAT, a non-parametric policy enforcement system that classifies user queries based on given categories of information that can change dynamically. The system operates by matching embedded user queries to synthetically generated queries in the database. This design enables low-cost policy updates through simple database modification, without retraining, while maintaining low latency. A comparison with strong prompt-based LLM baselines shows that Q-CAT delivers competitive performance in static environments, while offering practical advantages in dynamic policy settings, where policies change and the Query Pool can be updated from logs. Moreover, Q-CAT is up to two orders of magnitude faster than comparable-performance baselines in our experimental setup."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kang-etal-2026-q">
<titleInfo>
<title>Q-CAT: A Non-Parametric Query Filtering System for Prohibited Content in Dynamic Enterprise Environments</title>
</titleInfo>
<name type="personal">
<namePart type="given">Juheon</namePart>
<namePart type="family">Kang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Eunkyung</namePart>
<namePart type="family">Choi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hongseok</namePart>
<namePart type="family">Oh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hyunseo</namePart>
<namePart type="family">Shin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">JuYoun</namePart>
<namePart type="family">Son</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yunjin</namePart>
<namePart type="family">Bae</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Wonseok</namePart>
<namePart type="family">Hwang</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>Large Language Models (LLMs) are increasingly being adopted in enterprise environments, where employees may intentionally or accidentally submit prompts containing information that could cause legal or economic harm if leaked. Organizations may define categories of information (policies) that are prohibited from being submitted to commercial LLMs, but these definitions can vary between companies and evolve over time. This variability makes parametric classifiers less practical, as they require retraining whenever policies change. Here, we present Q-CAT, a non-parametric policy enforcement system that classifies user queries based on given categories of information that can change dynamically. The system operates by matching embedded user queries to synthetically generated queries in the database. This design enables low-cost policy updates through simple database modification, without retraining, while maintaining low latency. A comparison with strong prompt-based LLM baselines shows that Q-CAT delivers competitive performance in static environments, while offering practical advantages in dynamic policy settings, where policies change and the Query Pool can be updated from logs. Moreover, Q-CAT is up to two orders of magnitude faster than comparable-performance baselines in our experimental setup.</abstract>
<identifier type="citekey">kang-etal-2026-q</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.1221/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>27870</start>
<end>27891</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Q-CAT: A Non-Parametric Query Filtering System for Prohibited Content in Dynamic Enterprise Environments
%A Kang, Juheon
%A Choi, Eunkyung
%A Oh, Hongseok
%A Shin, Hyunseo
%A Son, JuYoun
%A Bae, Yunjin
%A Hwang, Wonseok
%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 kang-etal-2026-q
%X Large Language Models (LLMs) are increasingly being adopted in enterprise environments, where employees may intentionally or accidentally submit prompts containing information that could cause legal or economic harm if leaked. Organizations may define categories of information (policies) that are prohibited from being submitted to commercial LLMs, but these definitions can vary between companies and evolve over time. This variability makes parametric classifiers less practical, as they require retraining whenever policies change. Here, we present Q-CAT, a non-parametric policy enforcement system that classifies user queries based on given categories of information that can change dynamically. The system operates by matching embedded user queries to synthetically generated queries in the database. This design enables low-cost policy updates through simple database modification, without retraining, while maintaining low latency. A comparison with strong prompt-based LLM baselines shows that Q-CAT delivers competitive performance in static environments, while offering practical advantages in dynamic policy settings, where policies change and the Query Pool can be updated from logs. Moreover, Q-CAT is up to two orders of magnitude faster than comparable-performance baselines in our experimental setup.
%U https://aclanthology.org/2026.emnlp-main.1221/
%P 27870-27891
Markdown (Informal)
[Q-CAT: A Non-Parametric Query Filtering System for Prohibited Content in Dynamic Enterprise Environments](https://aclanthology.org/2026.emnlp-main.1221/) (Kang et al., EMNLP 2026)
ACL