Artificial Intelligence Policy

Overview

At-Taqaddum permits the use of Generative Artificial Intelligence (AI) and AI-assisted technologies only as a supporting tool, never as a substitute for human scholarship, and requires every such use to be disclosed. This policy applies to authors, reviewers, editors, and the publisher, the Institute for Quality Assurance (Lembaga Penjaminan Mutu/LPM), Universitas Islam Negeri Walisongo Semarang.

As a journal devoted to quality assurance, governance, and organizational quality in educational institutions, At-Taqaddum holds its own publication process to the same standard of quality it studies. Responsible AI use can improve language, efficiency, and scholarly communication. Undisclosed or inappropriate use, however, threatens originality, the validity of evidence, and academic accountability.

This policy follows the guidance of the Committee on Publication Ethics (COPE) and the values of Islamic scholarly ethics, namely amanah (trustworthiness), ṣidq (truthfulness), ‘adl (fairness), and itqān (excellence and meticulousness in work). All manuscripts published in At-Taqaddum must fundamentally reflect human intellectual contribution and scholarly judgment. The policy will be reviewed periodically as technologies and publishing practices evolve.

Foundational Principles

At-Taqaddum adopts six principles that govern every use of AI in its publication process.

  • Human accountability (mas’ūliyyah): Authors, reviewers, and editors bear full responsibility for the accuracy, validity, and originality of their work, regardless of the tools used.
  • Transparency (ṣidq): Any substantive use of AI must be disclosed clearly, completely, and honestly.
  • Supplement, not substitute: AI may assist, but it cannot replace human reasoning, creativity, scholarly judgment, or moral agency.
  • Confidentiality and data protection (amanah): Manuscripts, review reports, institutional documents, and research participants’ personal data are trusts that must not be exposed to external AI systems. All use must comply with Law of the Republic of Indonesia No. 27 of 2022 on Personal Data Protection.
  • Quality and verification (itqān): Every AI-assisted output must be checked by a human against original sources and verified data before it enters a manuscript, review, or editorial record.
  • Editorial oversight: The Editor-in-Chief and the publisher reserve the right to determine whether AI use in a submitted work is appropriate.

For Authors

Authors may use AI only for supportive, non-substantive tasks, and must disclose every use beyond basic proofreading.

Permitted Uses of AI

  • Copy-editing and proofreading of the author’s own text (grammar, syntax, spelling, and structure).
  • Translation of the author’s original work, provided the author verifies its accuracy and declares the use.
  • Creating explanatory diagrams or conceptual visualizations (for example, a quality-assurance cycle or an accreditation process flow) based on verified data, labeled “AI-generated”.
  • Creating graphs or charts from verified, existing data, provided they reflect the results accurately.
  • Ideation and reference discovery, used in the same way as a conventional search engine, with every source verified by the author.
  • Transcribing interviews or focus group discussions, provided participants have consented, no identifying data is exposed to tools that store or train on user input, and the author checks the transcript against the original recording.

Prohibited Uses of AI

  • Writing substantive content, including the abstract, introduction, literature review, methods, results, discussion, or conclusion.
  • Generating, simulating, manipulating, or fabricating data, survey responses, interview quotations, statistical outputs, figures, or tables.
  • Interpreting findings or coding qualitative data on the author’s behalf.
  • Producing literature reviews or summarizing prior research on the author’s behalf.
  • Generating or compiling references and citations, including any reference the author has not read and verified, or citing an AI tool as a primary source.
  • Generating or altering factual or evidential images, such as photographs, screenshots of systems, or scanned institutional documents.
  • Writing new analytical or statistical code.

Research Data and Confidential Institutional Documents

Research on quality assurance often involves accreditation evidence, internal audit reports, institutional performance data, and responses from students, lecturers, staff, alumni, or graduate users. Authors must not upload such material into AI tools that store, share, or train on user input unless it has been fully anonymized and the authors hold permission from the data owner. Authors remain responsible for protecting research participants and institutions in line with research ethics approval and Law No. 27 of 2022 on Personal Data Protection.

Disclosure Requirement

In line with the COPE position statement on authorship and AI tools, authors must disclose any substantive use of AI at the time of submission, naming the tool, its version, and how it was used.

  • Use in the research process (data collection, interview transcription, analysis support, or data visualization) must be described in the Methods section.
  • Use in manuscript preparation (language editing beyond basic proofreading, or translation) must be stated in an AI Declaration Statement placed before the References, using the format below.

“During the preparation of this manuscript, the authors used [AI tool name, version] for [specific purpose, e.g., language editing of the discussion section or translation from Indonesian]. The authors reviewed and verified all output and take full responsibility for the content of the publication.”

Basic proofreading with grammar or spelling checkers does not require disclosure. Disclosure exists to make the work transparent, not to penalize authors, so honest disclosure of permitted use will not count against a manuscript. Undisclosed AI use discovered during review or after publication is treated as misrepresentation under COPE guidance.

Author Responsibility

All authors remain fully accountable for:

  • The accuracy, validity, originality, and integrity of the entire manuscript, including any part produced with an AI tool, and any breach of publication ethics arising from it.
  • Ensuring that all AI-assisted content has been reviewed, verified, and edited by humans.
  • Upholding research ethics in line with Islamic principles, including honesty, transparency, and the avoidance of deception (tadlīs).

Authorship

In line with the COPE position statement on authorship and AI tools, AI tools and large language models cannot be listed as authors or co-authors, because they:

  • Cannot conceptualize research or exercise scholarly judgment.
  • Cannot take responsibility for the submitted work, declare the presence or absence of conflicts of interest, or manage copyright and licence agreements.
  • Cannot be held accountable for ethical or scientific integrity.

For Reviewers

Reviewers may use AI only to polish their own writing, and must never place any part of a manuscript into an AI tool.

Permitted Uses of AI

  • Improving the clarity, grammar, or readability of the reviewer’s own report, after the substantive evaluation is complete and without including manuscript content.
  • Checking general factual details (for example, the definition of a standard or a public regulation) without exposing manuscript content.

Prohibited Uses of AI

  • Uploading any part of a manuscript, its supplementary files, or its data into an AI tool. This breaches confidentiality and may violate data protection law.
  • Using AI to analyze, summarize, evaluate, or score a manuscript, or to draft the substance of a review.
  • Using AI to make or suggest a recommendation of acceptance, revision, or rejection.

Disclosure

Reviewers who use AI to refine the language of their report must state this briefly in their confidential comments to the editor, naming the tool and its purpose.

Reviewer Responsibility

Reviewers must:

  • Maintain strict confidentiality of all manuscripts and review materials.
  • Report any suspected AI misuse, such as fabricated data, non-existent references, or undisclosed AI-generated text, to the editor.
  • Exercise independent scholarly judgment in line with COPE and the Islamic ethical principles of fairness (‘adl) and trust (amanah).
  • Remain solely responsible for the intellectual content and conclusions of their review.

For Editors

Editors may use AI only for administrative, non-decisional tasks under human oversight; every editorial decision remains a human decision.

Permitted Uses of AI

  • Managing editorial workflows, schedules, and routine communications that contain no confidential manuscript content.
  • Running similarity and plagiarism checks through the journal’s approved services.
  • Matching manuscripts with potential reviewers, with the final choice made by the editor.
  • Improving language clarity in administrative correspondence.

Prohibited Uses of AI

  • Uploading manuscripts, review reports, or confidential correspondence into external AI systems.
  • Using AI to draft, summarize, or evaluate manuscripts, or to write decision letters on their substance.
  • Using AI to recommend or make decisions of acceptance, revision, or rejection.
  • Processing or storing confidential manuscripts or correspondence in external AI systems.
  • Replacing human ethical judgment or scholarly discernment.

Editorial Responsibility

Editors are responsible for:

  • Checking AI disclosures by authors and reviewers and following up on any that are incomplete.
  • Investigating suspected AI misuse on concrete evidence, such as fabricated references or data, in line with COPE guidance. AI-detection scores alone are not proof of misuse, and manuscripts must never be uploaded to public AI detectors.
  • Maintaining the integrity and confidentiality of the peer review and editorial process.

Publisher’s Use of AI

LPM UIN Walisongo Semarang, as publisher of At-Taqaddum, may use AI tools to optimize editorial workflows, such as plagiarism checks, metadata organization, and indexing support, on three conditions:

  • All AI use is subject to human oversight, checks, and monitoring.
  • No editorial decision relies solely on AI output.
  • Authors and reviewers are informed, on the journal website and in submission and review guidelines, which AI tools the journal uses and for what purpose.
  • Any AI-assisted language editing at the production stage must not change the meaning of the accepted manuscript and is returned to the authors for approval in the proofs.
  • All use complies with data privacy, copyright, and ethical obligations.

Ethical Commitment, Compliance, and Consequences

At-Taqaddum upholds ethical AI use consistent with COPE and Islamic ethics, emphasizing truthfulness (ṣidq) and accountability (mas’ūliyyah) in research, transparency in declaring AI use, and the protection of data privacy and intellectual property throughout the publication process.

Suspected violations are handled according to COPE flowcharts and the COPE retraction guidelines revised in 2025, which list undisclosed involvement of AI as a form of misrepresentation that warrants retraction. Those involved are given the opportunity to respond, and the journal may publish an expression of concern while a case is under investigation. Confirmed non-compliance may result in:

  • Authors: a request for correction and full disclosure, manuscript rejection, or retraction of a published article, with notification to the author’s institution where warranted.
  • Reviewers: removal from the review assignment and from the reviewer database, particularly for breaches of confidentiality.
  • Editors: corrective action by the publisher, including removal from editorial duties, for misusing AI or failing to enforce this policy.

AI Usage Reference Table

The table below summarizes, section by section, which AI uses are permitted; where it and the text above differ, the text above prevails.

Section

AI Usage Area

Example

Permitted?

Abstract

Abstract creation

Using AI to generate or expand the abstract from prompts or prior text

No

Abstract

Abstract copy-editing

Correcting grammar and clarity of the author’s own abstract

Yes

Introduction

Ideation

Exploring general ideas, background topics, or research directions, as with a search engine

Yes

Introduction

Introduction creation

Drafting or expanding the introduction with AI prompts

No

Introduction

Introduction copy-editing

Improving readability and grammar without altering meaning

Yes

Literature review

Literature discovery

Finding literature or identifying research gaps; every source verified by the author

Yes

Literature review

Literature review creation

Writing or summarizing prior research with AI

No

Literature review

Literature review copy-editing

Editing the language or format of the author’s own review

Yes

Methods

Methodology ideation

Exploring possible quality-assurance models, frameworks, or instruments for consideration

Yes

Methods

Methodology creation

Writing or generating the methods section with AI

No

Methods

Methodology copy-editing

Refining the author’s existing methods text for clarity and grammar

Yes

Data

Interview transcription

Transcribing recordings with participant consent and no exposure of identifying data; checked by the author

Yes (declare)

Data

Uploading confidential data

Placing identifiable survey responses, audit reports, or accreditation evidence into public AI tools

No

Results

Data generation

Creating, simulating, or completing data, responses, or statistical outputs

No

Results

Qualitative coding and interpretation

Having AI code interview data or interpret findings

No

Results

Data visualization

Making graphs or charts from verified, existing data that reflect results accurately

Yes (declare)

Results

Results copy-editing

Shortening or reformatting the author’s own written results for readability

Yes

Discussion

Discussion creation

Generating or expanding analytical or interpretive content

No

Discussion

Discussion copy-editing

Refining grammar, flow, or readability while preserving meaning

Yes

Conclusion

Conclusion creation

Generating or expanding the conclusion with AI prompts

No

Conclusion

Conclusion copy-editing

Editing the author’s conclusion for readability and clarity

Yes

Additional

Translation

Translating the author’s original, unpublished work, verified by the author

Yes (declare)

Additional

References and citations

Generating or automatically compiling reference lists or citations with AI

No

Additional

Acknowledgments and declarations

Polishing or formatting author-written statements

Yes

Code

Code generation

Writing new analytical or statistical code with AI prompts

No

Code

Code editing

Improving readability of, or commenting on, existing human-written code

Yes

Images and visuals

Conceptual diagrams

Creating process or conceptual diagrams, labeled “AI-generated” and checked for accuracy

Yes (declare)

Images and visuals

Factual or evidential images

Generating or altering photographs, system screenshots, or scanned documents used as evidence

No

Images and visuals

Artistic renderings

Producing decorative cover art or graphical abstracts with AI

No

Peer review

Review copy-editing

Reviewer refines the language of their own report, with disclosure to the editor

Yes (declare)

Peer review

Review content generation

Reviewer uses AI to evaluate, summarize, or judge a manuscript

No

Editorial

Workflow tools

Plagiarism checks, reviewer matching, or workflow management under human oversight

Yes

Editorial

Decision-making

Using AI to recommend acceptance or rejection

No