<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="dai-dong-2026-steal">
    <titleInfo>
        <title>Let Them Steal: Trapping Large Language Model Extraction Attacks with Knowledge Honeypot</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Yuyang</namePart>
        <namePart type="family">Dai</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Yushun</namePart>
        <namePart type="family">Dong</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 deployed as commercial APIs are vulnerable to model extraction attacks, while existing defenses either act too late or degrade utility for legitimate users. We propose Knowledge Trap, a defense that redirects extraction attacks toward low-transferability knowledge through a Honeypot Knowledge Graph (HKG) and breadcrumb-guided exploration. Instead of blocking queries or perturbing outputs, Knowledge Trap consumes the attacker’s limited query budget on knowledge with negligible downstream utility while preserving benign-user performance. Experiments in medical and financial domains show that Knowledge Trap reduces surrogate Agreement by 6.2% on average without degrading legitimate-user accuracy, outperforming existing defenses that impose measurable user impact. These results suggest that defending knowledge-space traversal is a practical direction for mitigating LLM extraction attacks.</abstract>
    <identifier type="citekey">dai-dong-2026-steal</identifier>
    <location>
        <url>https://aclanthology.org/2026.emnlp-main.411/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>9428</start>
            <end>9449</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Let Them Steal: Trapping Large Language Model Extraction Attacks with Knowledge Honeypot
%A Dai, Yuyang
%A Dong, Yushun
%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 dai-dong-2026-steal
%X Large language models deployed as commercial APIs are vulnerable to model extraction attacks, while existing defenses either act too late or degrade utility for legitimate users. We propose Knowledge Trap, a defense that redirects extraction attacks toward low-transferability knowledge through a Honeypot Knowledge Graph (HKG) and breadcrumb-guided exploration. Instead of blocking queries or perturbing outputs, Knowledge Trap consumes the attacker’s limited query budget on knowledge with negligible downstream utility while preserving benign-user performance. Experiments in medical and financial domains show that Knowledge Trap reduces surrogate Agreement by 6.2% on average without degrading legitimate-user accuracy, outperforming existing defenses that impose measurable user impact. These results suggest that defending knowledge-space traversal is a practical direction for mitigating LLM extraction attacks.
%U https://aclanthology.org/2026.emnlp-main.411/
%P 9428-9449

Markdown (Informal)

[Let Them Steal: Trapping Large Language Model Extraction Attacks with Knowledge Honeypot](https://aclanthology.org/2026.emnlp-main.411/) (Dai & Dong, EMNLP 2026)

• Let Them Steal: Trapping Large Language Model Extraction Attacks with Knowledge Honeypot (Dai & Dong, EMNLP 2026)

ACL

• Yuyang Dai and Yushun Dong. 2026. Let Them Steal: Trapping Large Language Model Extraction Attacks with Knowledge Honeypot. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 9428–9449, Budapest, Hungary. Association for Computational Linguistics.