Generative AI Policy

1. Purpose and Scope

The Journal of Pharmaceutical Chemistry (JPC) recognizes that generative artificial intelligence (AI), large language models (LLMs), AI-assisted writing applications, image-generation systems, and other AI-enabled technologies can support scientific research, manuscript preparation, data processing, and scholarly communication.

At the same time, the use of these technologies raises important concerns regarding scientific accuracy, originality, authorship, research integrity, confidentiality, intellectual property, data protection, and accountability.

This policy establishes the principles and requirements governing the responsible use of generative AI and AI-assisted technologies in the preparation, submission, peer review, editorial assessment, and publication of manuscripts in JPC.

The policy applies to all manuscript types considered by the journal, including original research articles, review articles, short communications, systematic reviews, methodological papers, and other scholarly contributions accepted by JPC.

The central principle is that AI may assist scientific work, but it cannot replace human scholarly responsibility, independent scientific judgment, or accountability for published research.

2. Definition of Generative AI and AI-Assisted Technologies

For the purposes of this policy, generative AI and AI-assisted technologies include software or computational systems that generate, transform, summarize, classify, interpret, or substantially modify content or information using artificial intelligence.

Examples include:

·         Large language models and conversational AI systems, such as ChatGPT, Claude, Gemini, and comparable tools.

·         AI-assisted writing, rewriting, paraphrasing, translation, and manuscript-editing applications.

·         AI-based image, illustration, diagram, graphical abstract, and figure-generation systems.

·         AI tools used for literature synthesis, research-question development, code generation, and scientific data interpretation.

·         AI-enabled systems used for computational modeling, cheminformatics, molecular design, bioinformatics, and other pharmaceutical chemistry research.

·         AI features integrated into scientific writing platforms, reference-management software, laboratory software, and data-analysis applications.

This policy distinguishes between AI used to prepare a manuscript and AI used as a component of the research methodology. Both may require disclosure, but the appropriate location and level of detail differ.

Routine digital tools that do not employ generative AI, such as ordinary word processors, conventional reference managers, and standard statistical software, are not automatically subject to the disclosure requirements for generative AI.

3. General Principles of Responsible AI Use

All individuals involved in the publication process must adhere to the following principles:

3.1 Human responsibility

Authors remain fully responsible for the accuracy, originality, validity, integrity, and ethical acceptability of every part of their submitted and published work, including material generated or modified using AI.

AI-generated outputs must be critically reviewed and independently verified before they are incorporated into a manuscript.

3.2 Transparency

The use of AI must be disclosed whenever required under this policy. Disclosures must identify the tool, explain its purpose, and describe its contribution to the manuscript or research process sufficiently for readers and editors to understand its role.

3.3 Scientific integrity

AI must not be used to fabricate, falsify, manipulate, or misrepresent research findings, experimental observations, analytical results, references, citations, or scientific evidence.

3.4 Originality and attribution

Authors must ensure that AI-assisted content does not reproduce protected or third-party material without appropriate attribution or permission. AI-generated content is not automatically original, copyright-free, or scientifically reliable.

3.5 Confidentiality and privacy

Confidential manuscripts, unpublished findings, personal data, and proprietary research information must be protected from unauthorized disclosure to AI systems or third parties.

3.6 Fairness and accountability

AI must not be used to discriminate against authors, manipulate peer review, misrepresent scientific contributions, or undermine impartial editorial decision-making.

4. Policy for Authors

4.1 Permitted uses of AI

JPC permits responsible use of AI-assisted tools, subject to this policy, applicable law, research ethics requirements, and the terms of the relevant software or service.

Permitted uses may include:

·         Correcting grammar, spelling, punctuation, and language errors.

·         Improving readability, clarity, sentence structure, and organization.

·         Assisting authors for whom English is not their first language.

·         Developing outlines, organizing sections, and improving the presentation of scientific arguments.

·         Supporting preliminary literature exploration and identifying potentially relevant research questions.

·         Assisting with code development, data-processing workflows, and computational tasks.

·         Supporting research design, molecular modeling, cheminformatics, bioinformatics, and statistical analysis when scientifically appropriate.

·         Creating explanatory illustrations or other visual materials when their use is permitted, their accuracy is verified, and all necessary disclosures and permissions are provided.

·         Generating preliminary summaries or extracting information from sources that the authors subsequently verify.

Permitted use does not relieve authors of responsibility for verifying all outputs or complying with the disclosure requirements below.

4.2 Uses that are prohibited

Authors must not use AI to:

1.      Fabricate experimental results, observations, datasets, spectra, chromatograms, microscopy images, clinical data, or other research evidence.

2.      Falsify, selectively alter, or deceptively manipulate data, images, figures, or analytical outputs.

3.      Generate fictitious references, nonexistent publications, false DOIs, inaccurate quotations, or misleading citations.

4.      Present AI-generated hypotheses, interpretations, text, images, or other material as independently verified scientific findings when they have not been validated.

5.      Conceal plagiarism, duplicate publication, inappropriate text recycling, or unauthorized reproduction of third-party material.

6.      Misrepresent the extent of human contribution or falsely claim that experiments, calculations, analyses, or literature reviews were performed when they were not.

7.      Generate deceptive images, graphical abstracts, or visual representations that materially misrepresent scientific evidence.

8.      Evade plagiarism-detection, research-integrity, or editorial screening procedures.

9.      Create fictitious authors, reviewers, affiliations, institutional approvals, ethics approvals, funding statements, or declarations.

10.  Submit AI-generated work without sufficient human review, verification, and intellectual contribution.

The journal will assess suspected violations in context. AI involvement alone does not establish misconduct; the nature of the conduct, the evidence, the applicable requirements, and the author's response must be considered.

4.3 Disclosure of AI use in manuscript preparation

Authors must disclose the use of generative AI or AI-assisted tools when those tools have contributed to drafting, rewriting, translating, substantially restructuring, generating, or otherwise materially modifying manuscript content.

The declaration should identify:

·         The name of the AI tool or service.

·         The provider or developer, where reasonably available.

·         The purpose for which the tool was used.

·         The nature and extent of its contribution.

·         Confirmation that the authors reviewed and verified the output and accept responsibility for the final manuscript.

Routine spelling, punctuation, or basic grammar corrections that do not materially alter content generally do not require a separate declaration. Authors should nevertheless disclose substantive AI-assisted editing or rewriting.

Where a tool is used for research design, data collection, computational analysis, image generation, or interpretation of scientific results, the use must also be described in the relevant Methods section or another appropriate section, with sufficient technical detail for evaluation and reproducibility.

Authors must disclose relevant AI use in the submission process and include the appropriate statement in the manuscript. JPC may request additional information where necessary.

4.4 Recommended disclosure statement

Authors who have used AI during manuscript preparation may use the following template:

“During the preparation of this manuscript, the author(s) used [name of AI tool and version, if available] for [specific purpose and nature of use]. The author(s) critically reviewed and verified the AI-assisted output, made the final editorial and scientific decisions, and take full responsibility for the accuracy, originality, integrity, and content of the published work.”

The statement must be adapted to reflect the actual use of AI. Authors must not use this declaration to imply that verification was performed when it was not.

If AI was used only for research methods or data analysis, the relevant details should be reported in the Methods section rather than relying solely on a general writing declaration.

If no declaration is required because the only use involved routine spelling, punctuation, or basic grammar correction, a separate statement is not necessary under this policy.

4.5 AI tools must not be listed as authors

AI systems, chatbots, language models, and other non-human AI tools cannot be listed as authors or co-authors of a JPC publication.

Authorship requires human accountability for the research, approval of the final manuscript, and the ability to address questions concerning the work. AI systems cannot independently meet these requirements.

All listed authors must satisfy the journal's authorship criteria and must approve the submitted and final versions of the manuscript.

AI assistance must not be used to misrepresent an author's qualifications, contribution, or responsibility.

4.6 Verification of scientific information and references

Authors must independently verify all AI-assisted scientific content, including:

·         Chemical names, molecular structures, formulas, reaction schemes, and stereochemical assignments.

·         Experimental procedures, reagent quantities, reaction conditions, yields, and characterization data.

·         Spectroscopic and analytical interpretations, including NMR, FTIR, mass spectrometry, and chromatographic results.

·         Molecular docking, molecular dynamics, density functional theory, QSAR, ADMET, and other computational findings.

·         Statistical calculations, model outputs, confidence intervals, and claims of significance.

·         Drug targets, biological mechanisms, pharmacological activities, toxicity assessments, and clinical claims.

·         Literature references, author names, journal titles, publication years, page ranges, and DOI information.

Authors must consult the original sources whenever possible. AI-generated citations or summaries must not be treated as substitutes for the original scientific literature.

All scientific claims must be supported by appropriate evidence, and any uncertainty or limitation must be reported honestly.

4.7 AI-generated images, figures, and graphical abstracts

AI-generated or AI-modified visual material must not misrepresent experimental evidence or imply that a generated image is a genuine experimental observation.

Authors must:

·         Disclose AI involvement in the creation or substantive modification of figures, illustrations, graphical abstracts, or other relevant visual content.

·         Identify the affected figure or material and explain the nature of the AI contribution.

·         Verify scientific labels, molecular structures, numerical annotations, legends, and other visual details.

·         Ensure that the tool's terms of use permit the intended publication and that third-party rights are respected.

·         Preserve original experimental images and data where applicable.

·         Provide original files, unprocessed images, or other supporting evidence when reasonably requested by the editors.

AI-generated illustrative artwork must be clearly distinguished from experimental photographs, microscopy images, diagnostic images, or other evidence-based scientific visuals.

AI must not be used to fabricate or deceptively modify experimental images, remove inconvenient observations, introduce nonexistent features, or alter results in a misleading manner.

Where the authenticity of a scientific image is central to a manuscript, the editors may request the original source files and supporting data.

4.8 AI use in research and computational methods

JPC recognizes that AI can be an integral part of modern pharmaceutical chemistry research. Authors may use appropriate AI or machine-learning methods for molecular property prediction, drug discovery, virtual screening, cheminformatics, bioinformatics, image analysis, and other legitimate research purposes.

When AI is used as part of the scientific methodology, the Methods section must provide sufficient information for critical evaluation and, where appropriate, reproducibility. Depending on the method, this may include:

·         The AI model, software, platform, version, and relevant configuration.

·         The purpose of the analysis and the research question addressed.

·         Data sources, inclusion and exclusion criteria, and preprocessing procedures.

·         Training, validation, and test-set construction, where applicable.

·         Relevant model parameters, prompts, computational settings, or decision criteria.

·         Validation procedures, performance metrics, uncertainty assessments, and limitations.

·         Human intervention, quality-control procedures, and any material post-processing.

Authors must not describe AI-assisted predictions as experimentally confirmed results unless appropriate experimental validation has been performed.

For computational drug discovery, predicted binding affinity, docking scores, molecular dynamics observations, and in silico ADMET predictions must be interpreted within the limitations of the methods used. Such outputs must not be presented as proof of clinical efficacy, safety, or therapeutic benefit without appropriate supporting evidence.

4.9 AI-generated data and synthetic datasets

AI-generated or synthetic datasets may be used when scientifically justified, methodologically appropriate, and ethically acceptable.

Authors must clearly distinguish synthetic data from experimentally measured, observed, or clinically collected data. The generation method, intended use, validation procedures, and limitations must be explained.

Synthetic data must never be represented as genuine experimental observations or used to create false evidence of experimental validation.

Where synthetic data are used to train or test a model, authors should describe the dataset construction and explain how data leakage, circular validation, and other relevant sources of bias were addressed.

4.10 Copyright, licensing, and third-party rights

Authors are responsible for ensuring that AI-assisted content complies with applicable copyright law, licensing terms, confidentiality obligations, and third-party intellectual property rights.

AI-generated material must not be assumed to be free of copyright restrictions merely because it was produced by an AI system.

Authors must obtain necessary permissions for third-party text, images, diagrams, datasets, and other materials. They must provide appropriate attribution wherever required.

Authors must also ensure that the use of an AI tool is compatible with the intended scholarly publication and that any relevant licensing restrictions are respected.

4.11 AI use in systematic reviews and evidence synthesis

AI may assist with preliminary literature searching, screening, extraction, categorization, and summarization when its use is methodologically appropriate.

However, authors remain responsible for the completeness and accuracy of the review. They must verify citations, extracted information, study classifications, and conclusions against the original sources.

For systematic reviews, scoping reviews, and other evidence-synthesis studies, material AI involvement in screening, data extraction, or synthesis must be reported in the Methods section. Authors should describe the workflow, verification procedures, and any limitations introduced by AI.

AI must not be used to invent studies, misclassify evidence, conceal contradictory findings, or create a misleading impression of the available literature.

5. Policy for Peer Reviewers

5.1 Confidentiality of submitted manuscripts

Manuscripts submitted to JPC are confidential scholarly communications. Reviewers must not upload or disclose a submitted manuscript, any portion of it, its figures, supporting information, unpublished data, or confidential editorial correspondence to an external generative AI system.

This prohibition applies unless JPC has explicitly authorized a specific workflow and the authors' confidentiality, intellectual property rights, and applicable data-protection requirements are adequately protected.

Reviewers must not assume that a tool is confidential merely because it is described as secure, private, or suitable for professional use.

5.2 Permitted supportive uses

Subject to confidentiality requirements and any instructions issued by JPC, reviewers may use AI-assisted tools for limited, non-confidential support, such as improving the grammar or organization of their own review report.

AI must not replace the reviewer's independent assessment of scientific quality, originality, methodological soundness, ethical acceptability, significance, or suitability for publication.

5.3 Prohibited reviewer practices

Reviewers must not:

·         Upload confidential manuscript content to unauthorized AI systems.

·         Delegate the scientific assessment or publication recommendation to AI.

·         Use AI-generated statements as a substitute for independent critical judgment.

·         Submit unverified AI-generated criticism, references, or technical claims.

·         Use AI to disclose or exploit unpublished findings for personal, commercial, or competitive advantage.

·         Use AI in a manner that compromises impartiality, confidentiality, or the integrity of peer review.

5.4 Disclosure by reviewers

Reviewers must disclose material AI assistance used to prepare a review report when required by JPC. The disclosure should identify the tool and explain its purpose.

Routine spelling or punctuation checks that do not materially affect the review need not ordinarily be declared.

Reviewers remain responsible for every statement, recommendation, and criticism submitted to the journal.

6. Policy for Editors and Editorial Board Members

6.1 Editorial accountability

Editorial decisions must be made by authorized human editors based on the manuscript's scientific merit, relevance, methodological quality, ethical compliance, peer-review reports, and the journal's published policies.

AI may provide limited administrative or analytical assistance, but it must not independently determine acceptance, rejection, reviewer selection, or allegations of misconduct.

6.2 Confidentiality and manuscript handling

Editors and editorial-board members must not upload confidential manuscripts, reviewer reports, author identities, unpublished findings, or editorial correspondence to unauthorized generative AI tools.

Any AI-assisted editorial workflow must be explicitly authorized by the publisher or journal, protect confidential information, comply with applicable data-protection requirements, and preserve human oversight.

6.3 Use of AI for editorial support

Where appropriately authorized, AI may assist with non-confidential administrative tasks, language improvement, workflow organization, or preliminary identification of relevant editorial issues.

Editors must independently verify AI-generated information and must not rely on AI alone to identify misconduct, assess scientific validity, or determine the outcome of a submission.

6.4 Editorial disclosure

If JPC uses AI tools in manuscript editing, production, or other material editorial activities, the journal or publisher should provide appropriate disclosure to authors and, where relevant, readers.

Editorial use must not compromise the confidentiality of manuscripts, the independence of peer review, or the accuracy of published content.

7. Disclosure Requirements and Placement

JPC expects authors to disclose material AI use clearly and in the appropriate location.

Type of AI use	Required disclosure
Substantive manuscript drafting, rewriting, or restructuring	AI declaration in the manuscript and disclosure during submission
Routine spelling, punctuation, or basic grammar correction only	Separate declaration generally not required
AI used in data collection, analysis, modeling, or interpretation	Methods section, with relevant technical details
AI-generated or materially AI-modified scientific figures	Figure-specific disclosure and, where applicable, general AI declaration
AI-generated synthetic data	Methods section, including generation and validation procedures
AI-assisted systematic-review screening or extraction	Methods section, including verification procedures
Material AI assistance in preparing a peer-review report	Disclosure to the journal when required by its reviewer instructions
Authorized material AI use in editorial or production processes	Appropriate disclosure by the journal or publisher, where relevant

If AI has been used in more than one way, the manuscript must contain all applicable disclosures. A single general statement may not be sufficient to explain a material role in the research methodology.

8. Editorial Assessment and Verification

JPC may assess disclosed or suspected AI use as part of its normal editorial and research-integrity procedures.

Editors may request, where relevant:

·         Clarification of the AI tool, version, or workflow.

·         Additional methodological details.

·         Original experimental data, image files, or computational outputs.

·         Source references supporting AI-assisted claims.

·         Documentation of validation and quality-control procedures.

·         An explanation of how the authors verified AI-generated material.

·         Corrections to an incomplete or inaccurate disclosure statement.

AI-detection software must not be treated as definitive proof that a manuscript was generated by AI or that an author has committed misconduct. Any concern must be evaluated using appropriate evidence, the manuscript's scientific content, and a fair opportunity for the authors to respond.

9. Non-Compliance and Publication Ethics

Failure to comply with this policy may result in editorial action proportionate to the nature and seriousness of the issue.

Depending on the circumstances, possible actions include:

1.      Requesting clarification or a complete AI-use declaration.

2.      Requiring correction or revision of the manuscript.

3.      Requesting supporting data, original files, or methodological documentation.

4.      Pausing editorial assessment while a concern is investigated.

5.      Rejecting a manuscript when serious concerns remain unresolved.

6.      Correcting a published article where a material disclosure or content issue requires correction.

7.      Retracting a published article when the evidence and applicable retraction criteria justify that action.

8.      Referring substantiated concerns to the relevant institution or responsible authority, where appropriate and consistent with applicable requirements.

JPC will consider the available evidence, the severity and impact of the conduct, the author's explanation, and applicable publication-ethics standards before taking action.

AI use itself is not automatically misconduct. Misrepresentation, fabrication, falsification, plagiarism, serious confidentiality breaches, or other substantive violations may constitute misconduct irrespective of whether AI was involved.

10. Responsibilities of Authors, Reviewers, and Editors

Authors

Authors must disclose AI use where required, verify all AI-assisted content, maintain the integrity of the research, protect confidential information, obtain necessary permissions, and accept full responsibility for their manuscripts.

Corresponding authors

The corresponding author must ensure that all co-authors are informed of material AI use, that the relevant declarations are complete and accurate, and that the submitted manuscript complies with this policy.

Reviewers

Reviewers must maintain confidentiality, exercise independent scientific judgment, disclose AI assistance when required, and remain accountable for their reports.

Editors and editorial-board members

Editors must safeguard confidential information, apply the policy fairly, maintain editorial independence, and ensure that AI does not replace human judgment in publication decisions.

Publisher

The publisher is responsible for supporting implementation of this policy, providing appropriate instructions to users of the journal, and reviewing the policy as technology, scholarly standards, and applicable requirements evolve.

11. Relationship with Other JPC Policies

This policy must be read alongside JPC's policies on publication ethics, authorship and contributorship, plagiarism, peer review, research ethics, conflicts of interest, corrections and retractions, copyright, and data availability.

Where an AI-related issue also involves suspected plagiarism, fabricated data, image manipulation, undisclosed conflicts of interest, privacy violations, or other ethical concerns, the relevant JPC policies will also apply.

If a conflict arises between this policy and applicable law or binding institutional or regulatory requirements, the matter will be handled in accordance with the applicable requirements and the journal's established procedures.

12. Policy Review and Updates

JPC recognizes that AI technologies, scientific practices, and publication-ethics standards continue to evolve.

The journal and publisher will review this policy periodically and may revise it when necessary to address new technologies, emerging risks, updated scholarly guidance, or changes in applicable law.

Any material revision should be communicated through the journal's official website and, where appropriate, through instructions to authors, reviewers, and editors.

The version applicable to a manuscript will ordinarily be the version in force at the relevant stage of submission or editorial assessment, subject to any transition arrangements formally announced by JPC.

13. Final Statement

The Journal of Pharmaceutical Chemistry (JPC) supports the responsible use of generative AI and AI-assisted technologies when they contribute to transparent, reproducible, and scientifically rigorous research.

The journal does not permit AI to replace human authorship, independent scientific judgment, ethical responsibility, or accountability for published findings.

Every author, reviewer, and editor is responsible for ensuring that AI is used transparently, confidentially, ethically, and in a manner that protects the reliability and integrity of the scientific record.