1. General Principles and Policy Scope

This policy applies to all submissions, revisions, peer-review activities, editorial assessments, publication preparation, and post-publication matters handled by AMAR.

1. Human accountability. Human authors, reviewers, and editors retain full responsibility for their respective scholarly contributions, assessments, recommendations, and decisions.
2. Transparency. Substantive use of generative AI or AI-assisted technologies must be disclosed accurately whenever required by this policy.
3. Research integrity. AI-generated or AI-assisted material must not introduce fabricated evidence, misleading analyses, false citations, plagiarism, or unsupported scientific claims.
4. Confidentiality. Manuscripts, unpublished findings, proprietary financial information, audit documentation, research data, and editorial records must be protected against unauthorized disclosure.
5. Independent peer review. AI-generated assessments cannot substitute for independent evaluations conducted by qualified human reviewers.
6. Editorial independence. Acceptance, revision, rejection, correction, and retraction decisions remain under authorized human editorial responsibility.

For this policy, AI-assisted technologies include generative language models, automated text or image-generation systems, AI coding assistants, and other tools capable of generating or materially transforming scholarly content. Routine conventional software functions are distinguished from substantive generative AI use.

2. AI Usage Policy for Authors

2.1. Authorship and Intellectual Responsibility

AI systems must not be named as authors or co-authors. They cannot satisfy scholarly authorship requirements, approve the final manuscript, disclose conflicts of interest, or accept responsibility for research integrity.

All listed human authors must make genuine scholarly contributions, approve the submitted work, and remain accountable for its originality, methods, evidence, interpretation, references, and conclusions.

2.2. Permitted Uses of AI

Authors may use AI-assisted tools for the following purposes, subject to appropriate verification and disclosure:

1. Language editing. Improving grammar, spelling, sentence clarity, readability, and academic English without changing the scientific meaning or creating unverified substantive claims.
2. Translation assistance. Translating an author-prepared draft into English, provided the authors independently verify terminology, methodological meaning, and the accuracy of the translated content.
3. Programming and technical assistance. Supporting the preparation, explanation, debugging, or improvement of statistical, computational, and visualization scripts, provided all code and outputs are independently tested and validated.
4. Research-related AI methods. Employing AI as an explicitly designed research instrument, analytical procedure, or object of investigation when methodologically justified, ethically permitted, transparently documented, and independently evaluated.
5. Explanatory graphics. Preparing non-deceptive explanatory diagrams or conceptual illustrations when their AI-assisted origin is disclosed where required, their accuracy is verified, and applicable third-party rights are respected.

2.3. Prohibited Uses of AI

Authors must not use AI to:

1. Delegate the central intellectual contribution, theoretical argument, literature synthesis, scientific interpretation, or conclusions to an AI system without genuine independent scholarly work and verification.
2. Fabricate or falsify empirical data, financial records, audit evidence, questionnaire responses, interview findings, statistical results, or research documentation.
3. Create synthetic audit evidence, financial statements, internal-control records, or investigative findings and misrepresent them as authentic observations.
4. Generate or alter research images, tables, charts, or figures in ways that distort actual evidence or misrepresent results.
5. Include invented or unverified references, quotations, journal articles, publication metadata, or DOI records.
6. Upload confidential corporate information, restricted datasets, identifiable research records, or third-party materials to unauthorized or insecure AI platforms.
7. Submit predominantly AI-generated scholarly content as an original intellectual contribution without appropriate human research, validation, and disclosure.

2.4. Verification of Sources and Findings

Authors must independently verify the existence, accuracy, and relevance of all cited sources. AI-generated bibliographic suggestions are not sufficient evidence that a publication exists or supports a scientific claim.

All in-text citations, reference entries, quotations, publication details, and DOI identifiers must be checked against authentic source records. References must follow the journal's required APA 7th edition style.

AI-assisted calculations, statistical estimates, visualizations, classifications, and analytical results must be checked against the actual research data and applicable methodology.

2.5. Research Papers and Literature Review Papers

The same integrity and disclosure requirements apply to both manuscript types accepted by AMAR: Research Paper and Literature Review Paper.

For Literature Review Papers, authors must independently verify the identification, eligibility, selection, evaluation, and interpretation of included publications. AI-generated summaries must not replace consultation of original scholarly sources.

Where AI supports a genuine research or review procedure, its role, configuration, relevant prompts or decision rules, verification procedures, and methodological limitations must be documented sufficiently for critical evaluation and reproducibility, where applicable.

3. AI Usage Policy for Editors

3.1. Human-Led Editorial Decisions

Scientific evaluation and editorial decisions must remain under the responsibility of qualified human editors. AI systems must not independently determine acceptance, rejection, revision requirements, research quality, or the scientific contribution of a manuscript.

3.2. Permitted Administrative Assistance

Subject to institutional authorization, confidentiality safeguards, and human oversight, editors may use appropriate automated tools to:

1. Check submission completeness, formatting, required declarations, and administrative compliance with journal instructions.
2. Support secure similarity and reference-integrity screening, with independent examination of all material findings.
3. Identify relevant fields of expertise and potential reviewer candidates from appropriately accessible scholarly information.
4. Improve the language and clarity of routine editorial correspondence and administrative communications.
5. Assist with structured internal checklists or administrative workflow organization without transferring confidential submissions to unauthorized services.

3.3. Editorial Restrictions

Editors must not:

1. Upload confidential manuscripts, extracts, review reports, author information, or editorial correspondence to public or insecure generative AI systems.
2. Use an AI-generated scientific assessment as a replacement for competent editorial evaluation or independent peer review.
3. Issue automatic rejection decisions solely because an AI detection tool reports a particular probability or numerical score.
4. Treat automatically generated comments, citations, methodological criticisms, or reviewer recommendations as verified scholarly findings without human examination.
5. Disclose confidential information to an external service without appropriate authorization and adequate safeguards.

3.4. Editorial Verification

Where automated screening indicates a possible problem, editors must examine the underlying evidence, assess its relevance, and request clarification when appropriate. Similarity percentages and AI-detection scores are screening indicators rather than definitive proof of misconduct.

Manuscripts undergoing peer review must continue to receive independent assessments from qualified human reviewers in accordance with AMAR's double-anonymous review policy.

4. AI Usage Policy for Peer Reviewers

4.1. Independent Scholarly Assessment

Peer reviewers are appointed for their disciplinary knowledge, methodological competence, and independent professional judgment. Every submitted review must reflect the reviewer's own evaluation of the manuscript.

4.2. Permitted Limited Assistance

Reviewers may use AI tools to improve grammar, spelling, clarity, or translation of comments they have independently prepared, provided that no confidential manuscript text, unpublished findings, identifiable information, or protected editorial content is disclosed to an unauthorized tool.

4.3. Prohibited Reviewer Practices

1. Uploading a submitted manuscript or its abstract, tables, figures, unpublished findings, or extracts to public or insecure AI services.
2. Requesting AI to generate a complete peer-review report or substitute for the reviewer's independent methodological and scholarly assessment.
3. Delegating the identification of scientific weaknesses, evaluation of validity, or acceptance recommendations to AI without independent scholarly evaluation.
4. Generating review comments based on invented references, unverifiable evidence, or content that the reviewer has not examined.
5. Using unpublished manuscript content for personal research, competing publications, commercial activities, or model training.

4.4. Reporting Suspected Misuse

Reviewers who identify credible indications of fabricated citations, falsified data, AI-related misrepresentation, or other research-integrity concerns must report specific evidence confidentially to the responsible editor.

Reviewers must not independently accuse authors, contact their institutions, or rely exclusively on AI-detection scores to establish misconduct.

5. Confidentiality, Data Security and Research Ethics

AMAR requires responsible protection of manuscripts, research records, proprietary business information, financial and audit documentation, participant information, and confidential editorial correspondence.

1. Access limitation. Confidential research and editorial materials may be accessed only for authorized purposes.
2. External AI systems. Confidential materials must not be disclosed to public or otherwise insecure AI services. Any proposed external processing involving confidential content requires prior explicit editorial authorization and verified safeguards.
3. Research permissions. Authors must respect institutional authorization, professional confidentiality, participant consent, contractual restrictions, and applicable data-protection requirements.
4. Authentic evidence. AI-generated or simulated financial records, audit evidence, transactions, or other documents must never be represented as authentic empirical observations.
5. Incident reporting. Suspected unauthorized disclosure or misuse must be reported promptly to the editorial office for appropriate assessment.

When AI-generated outputs are themselves the legitimate subject of a research study, authors must identify their synthetic nature, describe the procedures used, and distinguish them clearly from authentic empirical evidence.

6. Declaration and Disclosure of AI Use

6.1. When Disclosure Is Required

Routine spelling correction, conventional grammar checking, formatting, or straightforward language correction without substantive content generation does not normally require a separate AI declaration.

Substantive generative AI assistance must be disclosed, including extensive language restructuring, translation of a draft, generation or debugging of analytical code, AI-assisted research procedures, and other material assistance affecting preparation or analysis.

6.2. Where the Declaration Must Be Placed

No additional mandatory third submission file is introduced by this policy. Disclosure in the Title Page and appropriate methodological reporting are sufficient unless the editor requests justified supporting documentation.

6.3. Required Declaration Components

1. Tool identification: Name, developer, and model or version where identifiable and relevant.
2. Purpose and scope: The specific tasks, manuscript components, or research procedures assisted by AI.
3. Human verification: A statement explaining that all relevant outputs were examined, corrected, tested, and validated by the authors.
4. Responsibility: Confirmation that the human authors retain full responsibility for the accuracy, originality, and integrity of the submitted work.

6.4. Standard Declaration Templates

These templates must be adapted to the actual research process. An AI declaration does not legitimize a use that is otherwise prohibited by this policy.

7. Policy Violations, Investigation and Sanctions

7.1. Fair Investigation and Due Process

Suspected AI-related misconduct must be evaluated through an evidence-based and proportionate editorial procedure. A numerical AI-detection score alone does not establish a violation.

The editorial office may examine relevant manuscript content, bibliographic records, research documentation, methodological materials, disclosures, and explanations provided by the concerned party.

Before a sanction is imposed, the concerned party should receive a reasonable opportunity to respond, subject to confidentiality, conflict-of-interest safeguards, and applicable publication-ethics procedures.

7.2. Possible Actions Concerning Authors

1. Minor disclosure deficiencies. Authors may be required to amend or complete an omitted or inadequate AI-use declaration when the underlying assistance was permitted and the scholarly record remains reliable.
2. Materially deficient submissions. When verified misuse compromises originality, source integrity, scientific interpretation, or manuscript reliability, AMAR may require substantial correction or reject the manuscript.
3. Serious research misconduct. Established fabrication, falsification, deliberate misrepresentation, severe confidentiality violations, or fundamentally unreliable published findings may lead to retraction and other proportionate actions, including institutional referral or submission restrictions where justified.

7.3. Possible Actions Concerning Reviewers

A review generated without independent human assessment, or produced through a serious confidentiality breach, may be invalidated and excluded from the editorial record. Depending on the evidence and severity, the journal may issue a warning, terminate the review assignment, or remove the individual from its active reviewer pool.

7.4. Possible Actions Concerning Editors

Verified editorial misuse may result in reassignment of the manuscript, withdrawal of handling authority, formal corrective measures, or review of the individual's editorial appointment. Conflicted individuals must not determine the outcome of an investigation involving their own conduct.

7.5. Post-Publication Actions

If concerns arise after publication, AMAR may seek an author explanation, request additional documentation, consult independent experts, and publish a correction, Expression of Concern, or retraction where warranted.

Any formal post-publication notice must be transparent and linked to the relevant article. Published records must not be silently altered to conceal substantive research-integrity problems.

8. Responsibilities and Implementation

This policy complements AMAR's existing authorship, peer-review, research-ethics, plagiarism, confidentiality, complaints, and post-publication policies. It should be interpreted consistently with those requirements and with applicable publication-ethics standards.

The policy applies prospectively when formally published by the journal. Historical submissions and publications are not automatically invalidated by subsequent policy changes. Any concerns about earlier records must be assessed on their merits under the applicable research-integrity procedures.