The AI Paradox in L2 Writing: Why Helpful Feedback Creates Unhelpful Dependency in Higher Education

The AI Paradox in L2 Writing: Why Helpful Feedback Creates Unhelpful Dependency in Higher Education | 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 Systematic Review The AI Paradox in L2 Writing: Why Helpful Feedback Creates Unhelpful Dependency in Higher Education Mohamed SEDDIKI, Souhila KORICHI This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8731897/v2 This work is licensed under a CC BY 4.0 License Status: Posted Version 2 posted You are reading this latest preprint version Show more versions Abstract Large Language Models (LLMs) offer immediate pedagogical benefits in higher education L2 writing instruction, yet sustained reliance creates critical, underexplored risks to learner autonomy, metacognitive judgment, and linguistic identity. This Critical Interpretive Synthesis (CIS) of 47 peer-reviewed studies (2015–2025) identifies which patterns of AI-assisted interaction led to successful versus unsuccessful educational outcomes in higher education. Spanning pre-LLM and Generative AI (GenAI) eras, it addresses three knowledge gaps: how sustained reliance affects learner confidence and autonomy (psychological); how algorithmic approval reshapes communicative intentionality (cognitive); and how algorithmic norms systematically marginalize non-Western linguistic expression (ideological). The synthesis develops the AI Dependency Syndrome (ADS) framework, which maps four fundamental trade-offs in AI-mediated writing: fluency gains versus metacognitive erosion, anxiety reduction versus autonomous judgment, grammatical accuracy versus communicative intent, and improved essay quality versus voice authenticity. These trade-offs arise from three interconnected mechanisms: Loss of Confidence in Unaided Production, Algorithmic Approval Bias, and Internalization of AI Norms, which recursively interact to reshape learner conceptions of competence, authorship, and linguistic identity. The framework integrates self-efficacy, communicative competence, and identity theories to provide nine observable diagnostic indicators enabling educators to recognize emerging dependency patterns in real classrooms. By clarifying what distinguishes effective from ineffective AI‑assisted interactions in higher education L2 writing, this synthesis positions AI dependency not as incidental overuse, but as a systemic, preventable condition that demands intentional pedagogical attention. Critical Interpretive Synthesis (CIS) Higher Education Human-AI Interaction Large Language Models (LLMs) Learner Autonomy Linguistic Identity L2 Writing Instruction Metacognitive Development Trade-offs in AI-Mediated Learning Full Text Additional Declarations The authors declare no competing interests. Supplementary Files AppendixA.docx SupplementaryMaterialsA.docx Cite Share Download PDF Status: Posted Version 2 posted You are reading this latest preprint version Show more versions 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-8731897","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Systematic Review","associatedPublications":[],"authors":[{"id":583814267,"identity":"e58dd933-4d9d-42e9-bdde-8e39eb427fa1","order_by":0,"name":"Mohamed SEDDIKI","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABA0lEQVRIiWNgGAWjYHACAyBmZmzjYWx88ICBIQEkdAAoQpSWZoMEkrQ08DCwScC0MODTotvevIHxR4W1bB/P4baKhJraPAaJ3IOHeRis7XFpMTtzrICZ50y6cRtvY9uNhGPHixkk8hKAWtJxWmN2I8cA6JHDiW38jEAtbMcS9wNFDs5gOMyGU8v9NwaMP/9BtBQk/DuW2CAB0cKD2xYeAwbeBqAWoMMYEttqwFoOfGA4LIHbL2kFh3mOAf3Cc7BZIrHvQGIDzxugFoN0A5xajh/e+PBHjbXs/J70hx8+fKtLbGDPMf6QUIE7xEDgABL7MJTGaQcmqCNe6SgYBaNgFIwYAABid1zMRE6oSgAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0009-0001-3922-1535","institution":"University of Laghouat","correspondingAuthor":true,"prefix":"","firstName":"Mohamed","middleName":"","lastName":"SEDDIKI","suffix":""},{"id":583814268,"identity":"4fa62f1a-1f14-4554-aa67-46e6ba1865b4","order_by":1,"name":"Souhila KORICHI","email":"","orcid":"","institution":"University of Laghouat","correspondingAuthor":false,"prefix":"","firstName":"Souhila","middleName":"","lastName":"KORICHI","suffix":""}],"badges":[],"createdAt":"2026-01-29 13:08:52","currentVersionCode":2,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-8731897/v2","doiUrl":"https://doi.org/10.21203/rs.3.rs-8731897/v2","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103049557,"identity":"9f0af05d-36e8-4789-b58b-bc3d544f2abb","added_by":"auto","created_at":"2026-02-20 07:42:33","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":478724,"visible":true,"origin":"","legend":"","description":"","filename":"FinalManuscript1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8731897/v2_covered_f2081aed-4a60-4ffc-9472-ee765bb4c9db.pdf"},{"id":102950554,"identity":"b67aa8dc-5c7c-4e75-a159-39e7f07ac1c9","added_by":"auto","created_at":"2026-02-18 20:46:07","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":36300,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixA.docx","url":"https://assets-eu.researchsquare.com/files/rs-8731897/v2/a152527f07b35e75815f16e4.docx"},{"id":102950552,"identity":"ef7bff52-2ace-4e6e-a79e-1a9612881268","added_by":"auto","created_at":"2026-02-18 20:46:07","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":46040,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterialsA.docx","url":"https://assets-eu.researchsquare.com/files/rs-8731897/v2/e97d53cbfe83bf2622a9b3d1.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"The AI Paradox in L2 Writing: Why Helpful Feedback Creates Unhelpful Dependency in Higher Education","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"Critical Interpretive Synthesis (CIS), Higher Education, Human-AI Interaction, Large Language Models (LLMs), Learner Autonomy, Linguistic Identity, L2 Writing Instruction, Metacognitive Development, Trade-offs in AI-Mediated Learning","lastPublishedDoi":"10.21203/rs.3.rs-8731897/v2","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8731897/v2","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eLarge Language Models (LLMs) offer immediate pedagogical benefits in higher education L2 writing instruction, yet sustained reliance creates critical, underexplored risks to learner autonomy, metacognitive judgment, and linguistic identity. This Critical Interpretive Synthesis (CIS) of 47 peer-reviewed studies (2015–2025) identifies which patterns of AI-assisted interaction led to successful versus unsuccessful educational outcomes in higher education. Spanning pre-LLM and Generative AI (GenAI) eras, it addresses three knowledge gaps: how sustained reliance affects learner confidence and autonomy (psychological); how algorithmic approval reshapes communicative intentionality (cognitive); and how algorithmic norms systematically marginalize non-Western linguistic expression (ideological). The synthesis develops the AI Dependency Syndrome (ADS) framework, which maps four fundamental trade-offs in AI-mediated writing: fluency gains versus metacognitive erosion, anxiety reduction versus autonomous judgment, grammatical accuracy versus communicative intent, and improved essay quality versus voice authenticity. These trade-offs arise from three interconnected mechanisms: Loss of Confidence in Unaided Production, Algorithmic Approval Bias, and Internalization of AI Norms, which recursively interact to reshape learner conceptions of competence, authorship, and linguistic identity. The framework integrates self-efficacy, communicative competence, and identity theories to provide nine observable diagnostic indicators enabling educators to recognize emerging dependency patterns in real classrooms. By clarifying what distinguishes effective from ineffective AI‑assisted interactions in higher education L2 writing, this synthesis positions AI dependency not as incidental overuse, but as a systemic, preventable condition that demands intentional pedagogical attention.\u003c/p\u003e","manuscriptTitle":"The AI Paradox in L2 Writing: Why Helpful Feedback Creates Unhelpful Dependency in Higher Education","msid":"","msnumber":"","nonDraftVersions":[{"code":2,"date":"2026-02-18 20:46:02","doi":"10.21203/rs.3.rs-8731897/v2","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}},{"code":1,"date":"2026-02-01 15:14:03","doi":"10.21203/rs.3.rs-8731897/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":"fd53740a-a952-4a11-9ed9-c0db908fd7b9","owner":[],"postedDate":"February 18th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-03T12:57:27+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-18 20:46:02","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v2","identity":"rs-8731897","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8731897","identity":"rs-8731897","version":["v2"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}