<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="chernikov-etal-2026-pawn">
    <titleInfo>
        <title>PAWN++ in the Wild: Evaluating AI-Generated Text Detection on the Web</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Oleg</namePart>
        <namePart type="family">Chernikov</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Evgenii</namePart>
        <namePart type="family">Ptitsyn</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Albina</namePart>
        <namePart type="family">Burlova</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 13th Web-as-Corpus Workshop</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Pedro</namePart>
            <namePart type="family">Ortiz Suarez</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Laurie</namePart>
            <namePart type="family">Burchell</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Nikola</namePart>
            <namePart type="family">Ljubešić</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Yves</namePart>
            <namePart type="family">Scherrer</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Veronika</namePart>
            <namePart type="family">Laippala</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Vuk</namePart>
            <namePart type="family">Dinić</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Thom</namePart>
            <namePart type="family">Vaughan</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">979-8-89176-481-1</identifier>
    </relatedItem>
    <abstract>Machine-generated text is becoming a routine part of the web. Most detectors, however, are still evaluated on curated datasets rather than real web content. We introduce PAWN++, a compact supervised detector that learns which likelihood signals from two frozen language models are most informative. On MAGE, our method reaches 98.36 AUROC in distribution and 98.42 under a joint domain and generator shift. It also transfers well to external benchmarks without target-specific training.\par To test whether this robustness extends to the web, we construct WATTLE, a benchmark of 41,300 texts in which each machine output is tied to its original human web page. The resulting data capture realistic transformations of web content as well as variation across languages and time. We use WATTLE to compare our method with seven other detectors. Performance deteriorates sharply on realistic rewrites: our method reaches 70.1 AUROC overall and 58.1 on text produced by a frontier model. Likelihood-based detectors fall below chance, and no evaluated system identifies more than 12% of frontier rewrites at a 1% false-positive rate. These results expose a substantial gap between standard evaluation scores and reliable detection on the web.</abstract>
    <identifier type="citekey">chernikov-etal-2026-pawn</identifier>
    <location>
        <url>https://aclanthology.org/2026.wac-1.25/</url>
    </location>
    <part>
        <date>2026-10</date>
        <extent unit="page">
            <start>282</start>
            <end>296</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T PAWN++ in the Wild: Evaluating AI-Generated Text Detection on the Web
%A Chernikov, Oleg
%A Ptitsyn, Evgenii
%A Burlova, Albina
%Y Ortiz Suarez, Pedro
%Y Burchell, Laurie
%Y Ljubešić, Nikola
%Y Scherrer, Yves
%Y Laippala, Veronika
%Y Dinić, Vuk
%Y Vaughan, Thom
%S Proceedings of the 13th Web-as-Corpus Workshop
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-481-1
%F chernikov-etal-2026-pawn
%X Machine-generated text is becoming a routine part of the web. Most detectors, however, are still evaluated on curated datasets rather than real web content. We introduce PAWN++, a compact supervised detector that learns which likelihood signals from two frozen language models are most informative. On MAGE, our method reaches 98.36 AUROC in distribution and 98.42 under a joint domain and generator shift. It also transfers well to external benchmarks without target-specific training.\par To test whether this robustness extends to the web, we construct WATTLE, a benchmark of 41,300 texts in which each machine output is tied to its original human web page. The resulting data capture realistic transformations of web content as well as variation across languages and time. We use WATTLE to compare our method with seven other detectors. Performance deteriorates sharply on realistic rewrites: our method reaches 70.1 AUROC overall and 58.1 on text produced by a frontier model. Likelihood-based detectors fall below chance, and no evaluated system identifies more than 12% of frontier rewrites at a 1% false-positive rate. These results expose a substantial gap between standard evaluation scores and reliable detection on the web.
%U https://aclanthology.org/2026.wac-1.25/
%P 282-296

Markdown (Informal)

[PAWN++ in the Wild: Evaluating AI-Generated Text Detection on the Web](https://aclanthology.org/2026.wac-1.25/) (Chernikov et al., WAC 2026)

• PAWN++ in the Wild: Evaluating AI-Generated Text Detection on the Web (Chernikov et al., WAC 2026)

ACL

• Oleg Chernikov, Evgenii Ptitsyn, and Albina Burlova. 2026. PAWN++ in the Wild: Evaluating AI-Generated Text Detection on the Web. In Proceedings of the 13th Web-as-Corpus Workshop, pages 282–296, Budapest, Hungary. Association for Computational Linguistics.