Generative AI Policy

Generative AI Policy

Language, Technology, and Social Media (LTSM) recognises that generative artificial intelligence (GenAI) and AI-assisted technologies are increasingly used in research, academic writing, and scholarly communication. This policy establishes requirements for transparency, human accountability, confidentiality, research integrity, and responsible use of such technologies by authors, reviewers, and editors.

Generative AI may assist scholarly work, but it does not replace human intellectual responsibility and accountability. Authors remain fully responsible for the originality, accuracy, integrity, citations, interpretations, ethical compliance, and final content of every manuscript submitted to LTSM.

1. Scope of This Policy

This policy applies to the use of generative AI and AI-assisted technologies in activities associated with manuscript preparation, research reporting, editorial assessment, peer review, and publication.

It covers, where relevant:

• AI-assisted manuscript writing and language improvement;
• AI-assisted summarisation, restructuring, or rewriting;
• generative AI used to produce textual content;
• AI-generated or AI-modified images and illustrations;
• AI tools used as part of research methodology;
• AI-assisted data processing or analysis;
• use of AI by reviewers; and
• use of AI by editors and editorial staff.

Different requirements apply depending on whether AI is used merely to assist manuscript preparation or constitutes part of the research methodology itself.

2. Responsibilities of Authors

Authors who use generative AI or AI-assisted technologies remain responsible for all material submitted to the journal.

Authors must:

• review and verify AI-assisted or AI-generated content before submission;
• ensure that arguments, interpretations, and conclusions accurately represent the authors' scholarly judgment;
• verify all factual statements;
• verify every citation, reference, DOI, quotation, and attribution;
• ensure that confidential or sensitive information has not been improperly disclosed to an external AI service;
• ensure compliance with applicable copyright, privacy, data-protection, and research-ethics requirements;
• disclose relevant use of generative AI as required by this policy; and
• accept full responsibility for the final manuscript.

3. AI Cannot Be an Author

Generative AI systems, chatbots, large language models, and other AI tools cannot be listed as authors or co-authors of manuscripts published by LTSM.

Authorship requires responsibilities that can only be assumed by accountable human contributors, including:

• approval of the final manuscript;
• responsibility for the accuracy and integrity of the work;
• management of conflicts of interest;
• participation in correction of the scholarly record when necessary; and
• accountability for ethical and legal obligations associated with publication.

AI tools should therefore be disclosed where appropriate, but must not appear in the author list or be credited as an author.

4. Declaration of Generative AI Use

When generative AI or AI-assisted technologies have been used in manuscript preparation beyond routine spelling, grammar, reference management, or basic formatting assistance, authors must provide a transparent declaration.

The declaration should appear in a dedicated section entitled:

A suitable declaration is:

Where no declarable generative AI use occurred, authors may state:

LTSM has required an AI-use declaration for published articles since Vol. 2 No. 1 (June 2024).

5. Routine Tools That Normally Do Not Require Disclosure

Disclosure is normally not required for routine tools that perform limited technical functions without generating substantive scholarly content.

Examples may include:

• basic spelling correction;
• basic grammar checking;
• reference-management software;
• bibliographic formatting;
• routine word-processing functions; and
• basic formatting or accessibility assistance.

However, when a tool substantially generates, restructures, interprets, summarises, translates, or rewrites scholarly content, transparent disclosure is expected.

6. Unacceptable Uses of Generative AI

Generative AI must not be used in a way that fabricates, falsifies, misrepresents, or obscures the scholarly record.

Unacceptable practices include using AI to fabricate or knowingly distort:

• references, citations, or DOI information;
• research participants;
• interview quotations or qualitative evidence;
• research data;
• statistical or analytical results;
• experimental observations;
• ethical approval information;
• informed-consent records;
• author or reviewer identities;
• peer-review reports;
• figures or images represented as authentic empirical evidence; or
• other information material to the validity of the research.

Such practices may be investigated under the journal's Allegations of Research Misconduct Policy .

7. AI Used as Part of the Research Methodology

This policy does not prohibit the scholarly use of AI, machine learning, large language models, automated language-processing tools, or related technologies as part of legitimate research methodology.

Where AI constitutes part of the research method, authors should describe its use in the Methods section with sufficient detail for scholarly evaluation and reasonable reproducibility.

Where relevant, authors should report:

• the name of the model, system, software, or service;
• version or relevant access date;
• provider or developer;
• the purpose for which the technology was used;
• relevant prompts or procedural settings when methodologically important;
• input data or data-selection procedures;
• validation or verification procedures;
• human oversight;
• known methodological limitations; and
• other information necessary to understand how AI contributed to the research.

Methodological use of AI should be reported as part of the research method and should not be concealed within a general writing-assistance declaration.

8. AI-Generated and AI-Assisted Images

The acceptability of AI-generated or AI-assisted visual material depends on its purpose, transparency, accuracy, provenance, and role within the scholarly argument.

Legitimate computational image generation that forms part of the research methodology should instead be transparently described and evaluated as a methodological procedure.

9. Reviewers and Confidentiality

Manuscripts submitted to LTSM are confidential scholarly documents. Reviewers must not upload manuscripts or substantial manuscript content to external generative-AI services when confidentiality, intellectual-property rights, privacy, or control of the submitted material cannot be assured.

This requirement also applies to:

• figures and tables;
• datasets;
• supplementary material;
• unpublished findings;
• personally identifiable information; and
• confidential editorial correspondence.

Reviewers remain personally responsible for the content, quality, accuracy, professionalism, and integrity of their review reports.

AI tools must not replace the reviewer's independent scholarly judgment.

10. Editors and Editorial Staff

Editors and editorial staff must protect the confidentiality of submitted manuscripts and the integrity of the peer-review process.

Confidential manuscripts, reviewer reports, personal information, editorial correspondence, unpublished data, and other protected submission materials must not be uploaded to external generative-AI systems where confidentiality, privacy, intellectual-property rights, or data control cannot be assured.

Generative AI must not make independent editorial decisions concerning:

• desk rejection;
• reviewer selection;
• acceptance or rejection;
• research-integrity findings;
• ethical determinations; or
• post-publication actions.

Final editorial responsibility remains with qualified human editors.

11. Assessment of Suspected Undisclosed AI Use

LTSM does not treat the output of an automated AI-detection system as conclusive evidence that generative AI has or has not been used.

Where editors have reasonable concerns regarding undisclosed AI use, fabricated content, unreliable citations, authorship, or research integrity, authors may be asked to provide clarification or relevant supporting information.

Editorial decisions will be based on the available evidence, the nature of the concern, author responses, and the effect of the issue on the reliability and integrity of the manuscript.

12. AI-Assisted Translation

Authors may use AI-assisted translation tools, but remain responsible for ensuring that the translated manuscript accurately represents the intended scholarly meaning.

Particular care is required for:

• discipline-specific terminology;
• participant quotations;
• culturally specific concepts;
• legal or ethical terminology;
• interpretive qualitative data; and
• technical descriptions whose meaning could change through inaccurate translation.

Substantive use of generative AI for translation or rewriting should be disclosed where required under Section 4 of this policy.

13. Privacy, Personal Data, and Sensitive Information

Authors, reviewers, and editors must not disclose confidential or sensitive information to generative-AI systems in a manner inconsistent with informed consent, research ethics, confidentiality obligations, institutional requirements, data-protection rules, or applicable law.

Particular caution is required for participant data, unpublished datasets, personally identifiable information, confidential interviews, health or disability information, restricted corpora, private communications, and other sensitive research material.

14. Undisclosed or Inappropriate AI Use

Failure to disclose relevant AI use does not automatically establish research misconduct. The journal will evaluate the nature, extent, intention, and consequences of the undisclosed use.

Depending on the circumstances, LTSM may:

• request clarification;
• require a corrected disclosure statement;
• request revision of affected material;
• reject a manuscript where reliability or integrity is materially compromised;
• investigate suspected fabrication, falsification, plagiarism, or other misconduct; or
• take appropriate post-publication action when a significant issue is discovered after publication.

Post-publication action will follow the journal's Correction and Retraction Policy where applicable.

15. AI Research Submitted to LTSM

Research examining generative AI, large language models, automated speech processing, machine learning, human–AI interaction, or other AI-related technologies is welcome when it falls within the journal's Aims and Scope.

The presence of AI as a tool, platform, or research setting alone does not establish scope relevance. Manuscripts should make a substantive scholarly contribution to language, linguistics, communication, media, speech/hearing, human–computer interaction, or language-related technology.

16. Policy Review and Development

Generative AI technologies and their implications for scholarly publishing continue to evolve rapidly. LTSM may therefore review and update this policy as technologies, research practices, ethical standards, and scholarly-publishing guidance develop.

Authors, reviewers, and editors are expected to comply with the version of the policy applicable at the relevant stage of manuscript processing or publication.

Supporting Editorial Policies

17. Questions About Generative AI Use

Authors who are uncertain whether a particular use of generative AI requires disclosure are encouraged to contact the editorial office before publication.

LTSM supports responsible innovation in scholarly communication while maintaining the principle that human authors, reviewers, and editors remain accountable for the intellectual, ethical, and editorial decisions associated with published research.