@inproceedings{kassem-etal-2026-robust,
title = "How Robust Are Router-{LLM}s? Analysis of the Fragility of {LLM} Routing Capabilities",
author = {Kassem, Aly M. and
Sch{\"o}lkopf, Bernhard and
Jin, Zhijing},
editor = "Demberg, Vera and
Inui, Kentaro and
Marquez, Llu{\'i}s",
booktitle = "Proceedings of the 19th Conference of the {E}uropean Chapter of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
month = mar,
year = "2026",
address = "Rabat, Morocco",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.eacl-long.351/",
pages = "7496--7507",
ISBN = "979-8-89176-380-7",
abstract = "Large language model (LLM) routing has emerged as a crucial strategy for balancing computational costs with performance by dynamically assigning queries to the most appropriate model based on query complexity. Despite recent advances showing that preference-data-based routers can outperform traditional methods, current evaluation benchmarks remain limited{---}they largely focus on general model capabilities while overlooking task-specific behaviors and critical concerns such as privacy, safety, and potential backdoor vulnerabilities introduced through preference data. In response, we propose the DSC benchmark: Diverse, simple, and categorized, an evaluation framework that categorizes router performance across a broad spectrum of query types{---}including coding, translation, mathematics, human instructions, general knowledge, and LLM jailbreaking{---}and integrates privacy and safety assessments to reveal hidden risks. Our experiments on three preference-based routers and two commercial counterparts demonstrate that while these systems improve efficiency, they often make suboptimal, category-driven decisions; for instance, a BERT-based router directs all coding and mathematics queries to the most powerful LLM{---}even when simpler models would suffice{---}while routing jailbreaking attempts to weaker models, thereby elevating safety risks."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kassem-etal-2026-robust">
<titleInfo>
<title>How Robust Are Router-LLMs? Analysis of the Fragility of LLM Routing Capabilities</title>
</titleInfo>
<name type="personal">
<namePart type="given">Aly</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Kassem</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bernhard</namePart>
<namePart type="family">Schölkopf</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhijing</namePart>
<namePart type="family">Jin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-03</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Vera</namePart>
<namePart type="family">Demberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kentaro</namePart>
<namePart type="family">Inui</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lluís</namePart>
<namePart type="family">Marquez</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Rabat, Morocco</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-380-7</identifier>
</relatedItem>
<abstract>Large language model (LLM) routing has emerged as a crucial strategy for balancing computational costs with performance by dynamically assigning queries to the most appropriate model based on query complexity. Despite recent advances showing that preference-data-based routers can outperform traditional methods, current evaluation benchmarks remain limited—they largely focus on general model capabilities while overlooking task-specific behaviors and critical concerns such as privacy, safety, and potential backdoor vulnerabilities introduced through preference data. In response, we propose the DSC benchmark: Diverse, simple, and categorized, an evaluation framework that categorizes router performance across a broad spectrum of query types—including coding, translation, mathematics, human instructions, general knowledge, and LLM jailbreaking—and integrates privacy and safety assessments to reveal hidden risks. Our experiments on three preference-based routers and two commercial counterparts demonstrate that while these systems improve efficiency, they often make suboptimal, category-driven decisions; for instance, a BERT-based router directs all coding and mathematics queries to the most powerful LLM—even when simpler models would suffice—while routing jailbreaking attempts to weaker models, thereby elevating safety risks.</abstract>
<identifier type="citekey">kassem-etal-2026-robust</identifier>
<location>
<url>https://aclanthology.org/2026.eacl-long.351/</url>
</location>
<part>
<date>2026-03</date>
<extent unit="page">
<start>7496</start>
<end>7507</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T How Robust Are Router-LLMs? Analysis of the Fragility of LLM Routing Capabilities
%A Kassem, Aly M.
%A Schölkopf, Bernhard
%A Jin, Zhijing
%Y Demberg, Vera
%Y Inui, Kentaro
%Y Marquez, Lluís
%S Proceedings of the 19th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2026
%8 March
%I Association for Computational Linguistics
%C Rabat, Morocco
%@ 979-8-89176-380-7
%F kassem-etal-2026-robust
%X Large language model (LLM) routing has emerged as a crucial strategy for balancing computational costs with performance by dynamically assigning queries to the most appropriate model based on query complexity. Despite recent advances showing that preference-data-based routers can outperform traditional methods, current evaluation benchmarks remain limited—they largely focus on general model capabilities while overlooking task-specific behaviors and critical concerns such as privacy, safety, and potential backdoor vulnerabilities introduced through preference data. In response, we propose the DSC benchmark: Diverse, simple, and categorized, an evaluation framework that categorizes router performance across a broad spectrum of query types—including coding, translation, mathematics, human instructions, general knowledge, and LLM jailbreaking—and integrates privacy and safety assessments to reveal hidden risks. Our experiments on three preference-based routers and two commercial counterparts demonstrate that while these systems improve efficiency, they often make suboptimal, category-driven decisions; for instance, a BERT-based router directs all coding and mathematics queries to the most powerful LLM—even when simpler models would suffice—while routing jailbreaking attempts to weaker models, thereby elevating safety risks.
%U https://aclanthology.org/2026.eacl-long.351/
%P 7496-7507
Markdown (Informal)
[How Robust Are Router-LLMs? Analysis of the Fragility of LLM Routing Capabilities](https://aclanthology.org/2026.eacl-long.351/) (Kassem et al., EACL 2026)
ACL