AI Content Self-Detection for Transformer-based large Language Models

AI Content Self-Detection for Transformer-based large Language Models | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article AI Content Self-Detection for Transformer-based large Language Models Antonio Caiado, Michael Hahsler This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4640131/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The usage of generative artificial intelligence (AI)tools based on large language models, including ChatGPT, Bard, and Claude, for text generation, has many exciting applications with the potential for phenomenal productivity gains. One issue is authorship attribution when using AI tools. This is especially important in an academic setting where the inappropriate use regenerative AI tools may hinder student learning or stifle research by, creating a large amount of automatically generated derivative work. Existing plagiarism detection systems can trace the source of submitted text but are not yet equipped with methods to accurately detect AI-generated text. This paper introduces the dea of direct origin detection and evaluates whether generative AI systems can recognize their output and distinguish it from human-written texts. We argue why current transformer-based models may be able to self-detect their own generated text and perform a small empirical study using zero-shot learning to investigate if that is the case. Results reveal varying capabilities of AI systems to identify their generated text. Google’s Bard model exhibits the largest capability of self-detection with an accuracy of 94%, followed by OpenAI’s ChatGPT with 83%.On the other hand, Anthropic’s Claude model seems to be not able to self-detect. generative AI plagiarism paraphrasing origin detection Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4640131","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":332539985,"identity":"9457cf6b-88b9-4c28-bb36-7bc7564ecc58","order_by":0,"name":"Antonio Caiado","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtUlEQVRIiWNgGAWjYJCCAwxsDHIQJhsJWoxJ0wJSmdhAtBb+9sMPD/wos0mfP+2MAcOHssOEtUicSTM42HMuLXfD7RwDxhnniNDCcIPB4ABv2+HcDdI5BsxABmEd8jfYPxz823Y4XX42UMtfYrQY3OAxOAw0PIEB6DBmRmK0GJ7JKTgscy7NcMPttAKgp9IJa5E7fnzzxzdlNvLys5M3PvhRZk1YCwo4QKL6UTAKRsEoGAW4AADeIT3gJUbCMwAAAABJRU5ErkJggg==","orcid":"","institution":"Southern Methodist University","correspondingAuthor":true,"prefix":"","firstName":"Antonio","middleName":"","lastName":"Caiado","suffix":""},{"id":332539986,"identity":"c8949bc8-1b6c-4275-84de-1aa98b568d51","order_by":1,"name":"Michael Hahsler","email":"","orcid":"","institution":"Southern Methodist University","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"","lastName":"Hahsler","suffix":""}],"badges":[],"createdAt":"2024-06-26 05:45:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4640131/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4640131/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82760761,"identity":"5312f206-99b8-4943-b265-4916ade131cf","added_by":"auto","created_at":"2025-05-15 03:01:44","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":320665,"visible":true,"origin":"","legend":"","description":"","filename":"AIContentSelfDetectionforTransformerbasedLargeLanguageModels.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4640131/v1_covered_16a22901-ef16-4ad4-baf7-cd55fffe7be3.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"AI Content Self-Detection for Transformer-based large Language Models","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"generative AI, plagiarism, paraphrasing, origin detection","lastPublishedDoi":"10.21203/rs.3.rs-4640131/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4640131/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The usage of generative artificial intelligence (AI)tools based on large language models, including ChatGPT, Bard, and Claude, for text generation, has many exciting applications with the potential for phenomenal productivity gains. One issue is authorship attribution when using AI tools. This is especially important in an academic setting where the inappropriate use regenerative AI tools may hinder student learning or stifle research by, creating a large amount of automatically generated derivative work. Existing plagiarism detection systems can trace the source of submitted text but are not yet equipped with methods to accurately detect AI-generated text. This paper introduces the dea of direct origin detection and evaluates whether generative AI systems can recognize their output and distinguish it from human-written texts. We argue why current transformer-based models may be able to self-detect their own generated text and perform a small empirical study using zero-shot learning to investigate if that is the case. Results reveal varying capabilities of AI systems to identify their generated text. Google’s Bard model exhibits the largest capability of self-detection with an accuracy of 94%, followed by OpenAI’s ChatGPT with 83%.On the other hand, Anthropic’s Claude model seems to be not able to self-detect.","manuscriptTitle":"AI Content Self-Detection for Transformer-based large Language Models","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-07-31 05:45:19","doi":"10.21203/rs.3.rs-4640131/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"fa945c47-81d5-4d99-be42-42d6dab00959","owner":[],"postedDate":"July 31st, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-05-15T02:53:37+00:00","versionOfRecord":[],"versionCreatedAt":"2024-07-31 05:45:19","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4640131","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4640131","identity":"rs-4640131","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}