AI & Generative AI Policy
Human Accountability Remains Essential
AI tools may assist research or manuscript preparation where permitted, but they do not replace human responsibility for the accuracy, originality, integrity, ethical conduct, and final content of scholarly work.
1. Scope
This policy applies to AI and generative AI tools used in connection with research, manuscript preparation, submission, peer review, editorial assessment, and scholarly communication with PenScholar Journals — including large language models, generative text tools, image generation systems, AI-assisted analysis, AI coding tools, automated translation, and AI summarization tools. This is not an exhaustive list of every AI product.
2. General Principles
-
Human Responsibility
Humans remain responsible for scholarly work.
-
Transparency
Material AI use should be disclosed where required.
-
Accuracy
AI-generated claims and references must be independently verified.
-
Originality
AI must not be used to misappropriate or reproduce another person’s work.
-
Confidentiality
Confidential manuscripts and unpublished research must not be exposed to unauthorized third-party systems.
-
Research Integrity
AI must not be used to fabricate, falsify, manipulate, or misrepresent research.
3. Authors and AI
Authors may use AI tools for permitted assistance, but authors remain fully responsible for the submitted manuscript. AI output must not be treated as inherently accurate — before submission, authors must independently verify facts, citations, references, statistics, data interpretation, quotations, claims, figures, and conclusions.
4. AI-Assisted Writing
Lower-risk assistance — such as grammar correction, spelling, language refinement, translation assistance, and formatting help — is generally treated differently from more substantive uses, such as generating sections of a manuscript, analysis, interpretations, literature reviews, or conclusions. Not every grammar correction is treated as authorship-level AI use, but more substantive generation is more likely to require disclosure (see Section 9).
5. AI-Generated Content, Data & Analysis
Authors remain responsible for any AI-generated text included in a manuscript: verifying factual accuracy, checking for fabricated references and unintended plagiarism, and ensuring the content genuinely reflects the actual research. AI-generated material must never be used to conceal false claims.
Data and Synthetic Data
AI must not be used to fabricate research data and present it as genuine empirical evidence. This is distinct from legitimate synthetic data, which may be appropriate in some research designs where it is clearly identified, appropriately generated, and methodologically justified — synthetic data is not automatically prohibited, but it must be described transparently.
AI-Assisted Analysis
AI tools may be used for analysis where scientifically appropriate. Authors remain responsible for method selection, validation, interpretation, reproducibility, and reporting limitations, and should describe AI’s role sufficiently for readers to understand the research process where it materially affects methodology.
6. AI-Generated and AI-Modified Images
AI must not be used to fabricate experimental results, add or remove scientific evidence, manipulate research findings, or create misleading research figures. Where synthetic or AI-generated images are legitimately part of the research methodology, they must be clearly identified and appropriately described.
Scientific Image Processing
Ordinary image processing used in normal scholarly figure preparation is not restricted by this policy. Legitimate processing is acceptable where it does not misrepresent the underlying evidence; authors should disclose material processing where relevant.
7. AI, Citations and References
AI tools can generate non-existent references, incorrect citations, misquoted sources, and incorrect bibliographic information. Authors are responsible for checking every citation — an AI-generated reference must never be submitted without independent verification, and an AI tool’s citation should never be assumed authentic.
8. AI Is Not an Author
AI systems cannot take responsibility for scholarly work in the way human authors can, and are not listed as authors. Human authors remain responsible for accuracy, integrity, originality, ethical compliance, disclosure, and final approval of the manuscript.
Authors should remain able to identify and take responsibility for the work they submit — AI tools should not obscure who performed the research, who analyzed the data, who wrote the manuscript, or who approved the final version.
9. Disclosure of AI Use
Where material AI use is disclosed, useful information includes the name of the tool, its version where relevant, the purpose of use, where it was used in the manuscript, and the nature of the assistance. Unnecessary technical detail is not required, and trivial spelling correction does not require disclosure.
“An AI-assisted language tool was used to improve grammar and readability during manuscript preparation. The authors reviewed and approved all resulting text and remain responsible for the final manuscript.”
10. Reviewers and AI
Reviewers must protect manuscript confidentiality. Reviewers should not upload confidential manuscripts, unpublished research, proprietary data, or other protected material into external AI systems unless explicitly permitted and appropriately secured.
Reviewers remain responsible for review quality, accuracy, confidentiality, conflict disclosure, and independent scholarly judgment. AI must not replace a reviewer’s professional judgment, and an AI-generated review report must not be submitted without meaningful human evaluation.
11. Editors and AI
Editors may encounter AI tools in editorial workflows. Where AI-assisted tools are used, human oversight must remain, editorial confidentiality must be protected, decisions must remain accountable, potential bias should be considered, and sensitive manuscript information must be protected. PenScholar does not use an automated editorial decision system.
12. Confidentiality and AI Systems
Before using an external AI tool, consider whether it retains, processes, or reuses submitted information. Confidential material may include unpublished manuscripts, peer review reports, research data, proprietary information, personal information, and editorial correspondence. This principle applies regardless of which AI platform is used — PenScholar does not claim that any specific platform stores or trains on submitted data.
Authors, editors, and reviewers should not expose personal or sensitive information to AI systems without an appropriate legal and ethical basis.
AI Bias and Reliability
AI systems may produce biased outputs, incorrect claims, missing perspectives, stereotypes, or overconfident responses. AI output is not objective simply because it is computational, and authors must critically evaluate AI-generated material.
AI-Assisted Translation
AI translation may assist authors but does not transfer responsibility away from them. Authors should verify meaning, technical terminology, citations, context, and scholarly nuance in any AI-assisted translation.
13. Research Integrity and Misuse of AI
AI must not be used to fabricate data, falsify results, manipulate evidence, generate fake participants, generate fake citations, misrepresent research methods, or conceal misconduct.
Depending on the circumstances, inappropriate AI use — such as fabricated references, fake data, undisclosed substantive AI-generated content, misleading images, confidentiality breaches, plagiarism, or misrepresentation of authorship — may become an editorial or research integrity concern. The response depends on the facts and applicable policy; this policy does not prescribe an automatic penalty.
AI Detection
PenScholar does not claim to detect all AI-generated manuscripts and does not use an AI detection score as definitive proof of misconduct. Any automated signal, now or in the future, would be treated as an indication requiring human evaluation — not as conclusive evidence on its own.
Read Research Integrity Policy
14. Reporting AI-Related Concerns
Credible concerns about misuse of AI — such as fabricated references, AI-generated research data, misleading images, confidentiality breaches, or undisclosed substantive AI use — may be reviewed under PenScholar’s Research Integrity, Publication Ethics, Peer Review, Complaints & Appeals, and Corrections & Retractions policies as applicable. No fixed investigation process or specific outcome is promised in advance, and serious problems affecting the reliability of the scholarly record may require correction or other editorial action.
Related PenScholar Policies
Author Checklist Before Submitting
- AI-assisted content has been reviewed by the authors.
- Citations have been independently verified.
- AI has not fabricated research data.
- Figures and images accurately represent the research.
- Required AI use has been disclosed.
- Human authors take responsibility for the manuscript.
- Confidential or personal information has not been improperly exposed to external AI systems.