Generative AI and AI-Assisted Research Policy | Neutrosophic Optimization and Intelligent Systems

Purpose and Commitment to Research Integrity

Neutrosophic Optimization and Intelligent Systems (NOIS), published by Sciences Force LLC, recognizes the growing role of artificial intelligence in scientific research, computational analysis, software development, scholarly writing, and publishing.

The journal supports the responsible development and use of artificial intelligence when it advances legitimate scientific inquiry without compromising originality, accuracy, transparency, reproducibility, confidentiality, intellectual property, or human accountability.

This policy establishes the requirements governing generative AI and AI-assisted tools used by authors, editors, reviewers, guest editors, and authorized publishing personnel.

It distinguishes the use of artificial intelligence as a subject or methodology of scientific research from the use of generative tools to prepare, modify, evaluate, or publish scholarly manuscripts.

AI-assisted work is assessed according to the reliability of its scientific evidence and the integrity of its research practices, not simply according to whether artificial intelligence was used.

This policy is informed by the Committee on Publication Ethics (COPE), the publisher's Use of AI and Automated Tools Policy, and relevant recommendations of the International Association of Scientific, Technical and Medical Publishers (STM).

The journal maintains independent human scientific judgment throughout manuscript assessment and publication.

Scope and Definitions

This policy applies to artificial intelligence tools and automated systems that materially contribute to the creation, transformation, analysis, interpretation, evaluation, or presentation of scholarly content.

Relevant tools may include large language models, generative image and video systems, AI-powered coding assistants, machine learning services, automated statistical and data-analysis tools, AI-based translation systems, and research-support applications.

For the purposes of this policy:

Generative AI refers to systems capable of producing new text, images, code, audio, video, structured data, or other outputs in response to prompts or inputs.

AI-Assisted Manuscript Preparation refers to the use of AI tools to assist with drafting, language editing, translation, summarization, reference processing, figure preparation, or related writing activities.

AI-Assisted Research refers to the use of AI in research design, data processing, simulation, modeling, programming, statistical analysis, prediction, optimization, or scientific interpretation.

Editorial AI Tools refers to authorized systems used to support publishing administration, integrity screening, reference checking, or other editorial processes.

Material AI Use refers to assistance that substantively affects scholarly content, methodology, results, figures, code, interpretation, or the preparation of a manuscript and is therefore relevant to transparent assessment.

Conventional mathematical software, standard numerical solvers, ordinary statistical packages, and non-generative formatting utilities are governed primarily by the journal's methodological reporting and reproducibility requirements. They must be identified when relevant to reproducing the research, but their routine use does not automatically constitute generative AI-assisted manuscript preparation.

Human Authorship and Accountability

Artificial intelligence systems cannot be listed as authors or co-authors.

Authorship requires human intellectual responsibility, approval of the submitted work, accountability for scientific claims, and the ability to respond to questions about accuracy, originality, ethics, and publication rights.

AI systems cannot independently fulfill these obligations.

All listed authors remain responsible for the accuracy, integrity, originality, and completeness of their manuscript, including material generated, translated, revised, calculated, or otherwise assisted by AI.

The use of an AI tool does not transfer scholarly responsibility to the tool developer, service provider, software platform, or model.

Authors must review and verify all AI-assisted content before submission.

An AI system must not be included in the author list, contributor metadata, ORCID records, or corresponding author field.

Its use must instead be disclosed as required by this policy.

Human contributors who provide substantive scholarly contributions remain subject to the journal's normal authorship and contributorship requirements.

Permitted Use for Language Editing and Manuscript Preparation

NOIS permits responsible use of generative AI to support readability, grammar, language clarity, and translation under human supervision.

Authors may use such tools to improve the expression of scholarly material that they have developed and verified themselves.

Automated assistance must not introduce unsupported claims, alter the scientific interpretation, change mathematical meaning, or fabricate information.

Authors must review all suggested changes and retain responsibility for the final text.

Language assistance must not be used to misrepresent a manuscript as containing original reasoning, analysis, or findings that the authors have not produced and validated.

Routine non-generative spelling checks, formatting functions, and similar mechanical corrections do not ordinarily require a separate AI Use Statement.

Where a generative AI system has been used for language editing or translation, authors must disclose the assistance in accordance with the publisher's reporting requirements.

The preparation of standard administrative wording with AI assistance must not be used to conceal authorship contributions or substitute generated text for independently developed scientific content.

AI-Generated Scientific Content

Authors must not use generative AI as an unverified substitute for original scientific reasoning, mathematical proof, experimental analysis, literature assessment, or scholarly interpretation.

AI-generated hypotheses, explanations, proposed models, derivations, and research ideas may be investigated as preliminary material, but they must not be presented as established scientific findings without appropriate independent verification.

A manuscript's central scientific contribution must be attributable to the authors' research and supported by evidence that can be assessed independently.

Generating extensive scientific narratives, mathematical claims, literature reviews, or results with AI and presenting them as independently established scholarship without adequate verification and disclosure is unacceptable.

Authors must distinguish material verified through their own research from suggestions or outputs produced by an AI system.

The journal evaluates the scientific validity and provenance of the work rather than treating polished writing as evidence of scientific reliability.

Mandatory Disclosure of Generative AI Use

Authors who use generative AI or other materially relevant AI-assisted tools in manuscript preparation must disclose that use accurately and transparently.

The declaration must identify:

- The tool or service used, including its name and version where known.

- The purpose of the assistance.

- The manuscript components or research activities affected.

- The nature and extent of the AI contribution.

- The human verification and editing performed.

- Any material use involving code, figures, data, or numerical analysis.

- Whether confidential, unpublished, restricted, or identifiable information was supplied to the tool and, where applicable, the safeguards and authorization relied upon.

The declaration must describe the actual use rather than rely on vague assurances of responsible practice.

An AI Use Statement must be included in the manuscript when disclosure is required.

Use of AI in research methods must additionally be reported in the appropriate scientific methods section, with sufficient detail to evaluate the work.

Authors are not required to declare the absence of AI use when no AI tool was used.

A declaration that no AI was used must not be provided when relevant use did occur.

Disclosure does not, by itself, establish that an AI-assisted method is scientifically acceptable.

Placement of AI Disclosures

AI-assisted language editing, translation, or other manuscript-preparation assistance must be disclosed in a clearly labeled Generative AI and AI-Assisted Tools Disclosure statement.

This statement should appear within the manuscript's Statements and Declarations section or another appropriate location designated by the journal.

If AI was used as part of the scientific methodology, the relevant details must appear in the Methods, Computational Framework, Experimental Setup, or equivalent section.

The author declaration must be consistent with the technical description.

For double-anonymous review, the statement must not unnecessarily identify authors or disclose protected institutional information in files supplied to reviewers.

The final publication must contain the approved disclosure where applicable.

The journal may request clarification when an AI Use Statement does not adequately describe the contribution.

Artificial Intelligence as a Research Methodology

NOIS welcomes original research involving artificial intelligence, machine learning, deep learning, generative modeling, computational intelligence, and related techniques when the work falls within the journal's scientific scope.

The use of AI as a research methodology is legitimate when its design, implementation, evaluation, and scientific contribution are appropriately documented.

Authors must distinguish between established AI methods applied in research and external generative services used to create manuscript content.

Where AI forms a central part of the research, the manuscript must provide sufficient information about the models, data, algorithms, assumptions, computational environment, and evaluation methods.

A study proposing a new AI model is not disqualified merely because it relies on artificial intelligence.

Nevertheless, the work must demonstrate scientific rigor, adequate validation, and a meaningful contribution beyond routine use of a pre-existing system.

Mathematical Reasoning, Theorems, and Proofs

Mathematical claims produced or assisted by AI must undergo independent human verification.

Authors must not describe a proposition, theorem, lemma, or corollary as proven solely because an AI model generated a plausible argument.

The validity of a proof depends on its logical correctness and the satisfaction of the stated assumptions.

Authors must verify each material derivation, implication, transformation, and use of an external mathematical result.

AI-generated mathematical examples or counterexamples must be checked against the stated definitions and conditions.

Computer-assisted or automated theorem proving may be used legitimately when the verification process is documented and scientifically appropriate.

Where a formal proof assistant, symbolic computation system, or automated reasoning tool is essential to the result, the manuscript must describe its role and identify relevant software, versions, assumptions, and verification procedures.

Evidence from numerical examples must not be represented as a general mathematical proof.

Neutrosophic and Uncertainty-Based Research

Research involving neutrosophic structures, fuzzy systems, indeterminacy, uncertainty models, or associated mathematical operators must maintain precise definitions and valid mathematical relationships.

AI-generated definitions or formulas must not be accepted without examining their compatibility with the relevant framework.

Authors must verify the permissible domains of membership or uncertainty components, mathematical constraints, transformation rules, and other essential assumptions.

Claims concerning consistency, generalization, inclusion, equivalence, stability, monotonicity, or other mathematical properties must be independently justified.

AI-assisted generation of new terminology or mathematical notation does not, by itself, constitute scientific innovation.

Where AI is used to explore mathematical structures or discover candidate relationships, the manuscript must distinguish exploratory computational findings from rigorously established results.

The journal may request mathematical derivations, computational scripts, or other supporting materials necessary to assess such claims.

AI-Assisted Software Development and Code Generation

Authors may use AI coding assistants to support legitimate programming tasks, provided that the generated or modified code is reviewed, tested, and documented.

Human authors remain responsible for the correctness, security, originality, licensing, and scientific suitability of the code used in the research.

Where code materially affects the reported findings, authors must verify its implementation against the described algorithms and mathematical procedures.

Appropriate verification may include unit tests, reference calculations, comparison with established implementations, code inspection, or reproducible test cases.

Authors should document relevant software versions, dependencies, computational settings, and procedures required to obtain the principal outputs.

AI-generated code must not be represented as independently validated merely because it executes without errors.

Potential vulnerabilities, licensing restrictions, copied code, and dependencies introduced by automated tools must be considered.

Where material AI-assisted code cannot be shared, the authors must describe the legitimate restrictions and provide sufficient methodological information for scientific assessment.

AI-Assisted Statistical Analysis and Numerical Results

AI-generated statistical conclusions, numerical calculations, and analytical interpretations must be independently checked.

Authors must verify the suitability of the statistical tests, assumptions, formulas, input data, computational procedures, and reported outputs.

A language model's numerical answer is not sufficient evidence of mathematical or statistical correctness.

Where AI generates analysis scripts, authors should run and test those scripts using appropriate computational tools.

Reported tables, figures, statistical estimates, p-values, confidence intervals, optimization results, and other numerical outputs must agree with the verified analyses.

Material assumptions, data transformations, parameter settings, and sources of randomness must be documented.

Results must not be selectively modified or reported to exaggerate the performance of a proposed method.

Where independent reproduction is feasible, authors should provide suitable code, data, and instructions through the journal's research reproducibility procedures.

AI-Assisted Optimization and Computational Experiments

AI-assisted development of optimization algorithms, experimental protocols, benchmark implementations, or parameter-selection procedures must be described accurately.

Authors must ensure that the implemented method agrees with its mathematical formulation and reported pseudocode.

Computational claims must be supported by suitable evaluation procedures, including relevant baselines and comparable resource budgets.

Where AI is used to tune parameters, select features, generate benchmark instances, or optimize experimental configurations, the procedures must be disclosed when material to the results.

Authors must avoid allowing information from final evaluation data to influence the design or tuning of a method in ways that invalidate the comparison.

Stochastic experiments must account appropriately for variability and reproducibility.

AI assistance does not excuse inadequate baselines, unsupported convergence claims, misleading benchmark results, or incomplete computational documentation.

AI Models, Training Data, and Evaluation Transparency

Studies developing, training, fine-tuning, or evaluating AI models must document the essential components of the research methodology.

Where relevant, authors must identify the data sources, provenance, dataset composition, preprocessing procedures, model architecture, training strategy, evaluation design, and relevant computational settings.

The manuscript should describe the use of pretrained models, external services, model checkpoints, and other material research resources.

Authors must disclose relevant limitations involving training data rights, privacy, consent, sampling bias, performance variability, and generalizability.

Train, validation, and test partitions must be managed appropriately to prevent misleading evaluation through data leakage.

Where external generative services are used, authors should identify the service and available version information, dates of access, material parameters, and relevant prompt or retrieval procedures.

When model behavior cannot be reproduced exactly because of changes to third-party services or nondeterministic computations, the limitations must be explained.

Claims about AI reliability, robustness, safety, fairness, explainability, or practical effectiveness must be supported by appropriate evidence rather than promotional descriptions.

Prompting, External Models, and Reproducible Research

When prompts or other AI inputs materially affect research outcomes, authors must describe them with enough detail to support evaluation.

Relevant information may include prompt templates, task instructions, sampling parameters, model settings, retrieval resources, response-selection procedures, and human post-processing.

Where feasible and lawful, material prompts, configuration files, scripts, and output-processing procedures should be deposited with other research materials.

Authors are not required to disclose confidential information, personal data, or protected material merely to publish prompt documentation.

Redacted, generalized, or otherwise appropriately protected descriptions may be used when necessary, provided that they remain scientifically informative.

If a proprietary service prevents exact reproducibility, the manuscript must explain the limitation and describe reasonable verification or sensitivity analyses.

Prompts and generated responses must not be selectively reported to misrepresent the model's performance.

Scientific Images and Research Data Integrity

AI-based image generation, enhancement, reconstruction, segmentation, and other processing must not compromise the authenticity of scientific evidence.

Authors must not use generative AI to fabricate or falsify research observations, medical images, microscopy results, experimental measurements, benchmark outputs, or other material represented as genuine evidence.

AI-generated additions, deletions, inpainting, or transformations that create, remove, conceal, or materially alter features used as scientific evidence are prohibited when they misrepresent the underlying observations.

Legitimate image-analysis and computational reconstruction methods may be used when they are part of the disclosed research methodology and their limitations are adequately evaluated.

Authors must preserve relevant original files and document material processing operations.

Where necessary, the journal may request unprocessed source images, processing scripts, image metadata, or other supporting evidence.

An enhanced image must not be presented as an unprocessed observation.

The central requirement is that figures accurately represent the scientific data and do not create misleading evidence.

Illustrative AI-Generated Figures and Media

AI-generated conceptual diagrams and other clearly illustrative materials may be considered when they do not purport to represent observed scientific data.

Such materials must be identified accurately when their generative origin is relevant to interpretation.

Authors must ensure that the illustration does not misrepresent an experimental result, mathematical property, simulation, or empirical observation.

Illustrative AI-generated figures must not be substituted for actual graphs, measured data, statistical outputs, or reproducible scientific visualizations.

Authors remain responsible for copyright compliance, necessary permissions, appropriate attribution, and disclosure of relevant generation or editing methods.

The journal may request replacement of a figure when its origin, scientific meaning, or intellectual property status cannot be established adequately.

Synthetic Data, Simulated Observations, and AI-Generated Content

Synthetic data and simulated observations can be legitimate research resources when their nature and generation procedures are disclosed.

Authors must distinguish synthetic material from independently collected real-world observations.

The manuscript must explain the data-generation method, relevant assumptions, model settings, validation procedures, and limitations.

AI-generated data must not be described as empirical measurements unless it represents a clearly identified transformation of genuine measurements and the representation is scientifically justified.

Synthetic datasets must not be used to conceal missing real-world evidence or inflate sample sizes improperly.

Performance established on synthetic data must not be presented as proof of real-world generalizability without appropriate independent validation.

Where generated content is used as a research stimulus or evaluation item, the manuscript must explain its role and the safeguards adopted to prevent misleading conclusions.

Literature Searches, Citations, and References

AI tools may support preliminary exploration of research topics, terminology, and potential literature.

However, generated bibliographic details, quotations, summaries, and scientific claims must be checked against the underlying sources.

Authors must verify that cited publications exist and that their titles, author names, dates, journal information, DOIs, and relevant conclusions are accurate.

Fabricated references, misattributed findings, nonexistent quotations, and unsupported claims derived from AI-generated summaries are unacceptable.

Authors must not cite an AI chatbot response as a substitute for the scientific source on which a factual claim depends.

When an AI model or software platform is itself an object of research, it may be appropriately identified through technical documentation, software citations, model documentation, or other legitimate sources.

Relevant retracted publications must be identified and treated according to their actual status.

AI-assisted citation suggestions must not be used to manipulate citation metrics or introduce irrelevant references.

Confidentiality, Privacy, and Data Protection

Authors must respect applicable privacy, confidentiality, intellectual property, contractual, and research ethics requirements when using AI services.

Confidential manuscripts, unpublished research materials, identifiable research participant information, and restricted datasets must not be submitted to external AI systems without an appropriate legal and ethical basis and adequate safeguards.

Authors must consider whether the service stores inputs, retains generated outputs, uses submitted information for training, transfers information internationally, or permits access by unauthorized parties.

De-identification does not automatically eliminate privacy risks when individuals can reasonably be re-identified.

Where AI processing involves sensitive research data, authors must comply with applicable institutional approvals, consent requirements, data-use agreements, and security controls.

A generic assertion that an AI service is secure does not replace the need to assess its actual processing arrangements.

The journal may request clarification where AI-assisted processing creates a material confidentiality or data-protection concern.

Intellectual Property and Third-Party Rights

The use of generative AI does not remove authors' obligations concerning copyright, software licensing, attribution, confidential information, or other intellectual property rights.

Authors must assess whether AI-assisted text, images, code, or other materials may incorporate protected third-party content.

AI-generated outputs must not be presented as unquestionably free of intellectual property restrictions.

Where a tool materially assists in producing research content, the authors remain responsible for obtaining any permissions necessary for lawful publication.

The Creative Commons Attribution 4.0 International License used by NOIS applies to material that the relevant rights holders have authority to license.

It does not automatically establish ownership of AI-generated material or authorize the redistribution of third-party content.

Any licensing or ownership limitations relevant to the publication must be identified before acceptance.

Use of Generative AI by Peer Reviewers

To preserve research confidentiality and independent expert assessment, NOIS does not permit reviewers to use generative AI or AI-assisted content-generation systems to perform peer review of confidential submissions.

Reviewers must not upload manuscripts, supplementary data, unpublished figures, reviewer reports, editorial correspondence, or confidential author information to generative AI services.

Reviewers must not use generative AI to generate substantive review reports, assess mathematical proofs, evaluate scientific claims, or formulate recommendations concerning manuscripts under review.

This restriction applies even when a tool is described as private, unless NOIS formally adopts and publicly communicates a revised review policy after appropriate technical, legal, and ethical assessment.

Routine non-generative writing and accessibility tools that do not disclose confidential manuscript content or replace scientific judgment are not prohibited by this provision.

Reviewers remain responsible for their own critical scientific assessment and for complying with the journal's confidentiality and conflict-of-interest requirements.

Suspected unauthorized AI use must be reported confidentially to the handling editor.

Use of AI by Editors and Guest Editors

Editorial decisions must be made by authorized human editors.

Generative AI must not independently determine manuscript acceptance, rejection, reviewer selection, scientific validity, or research misconduct findings.

Editors and Guest Editors must not submit confidential manuscripts, reviewer identities, private reports, or unpublished research information to unauthorized external AI tools.

NOIS does not permit generative AI to replace editorial reading, specialist judgment, independent review, or substantive scientific decision-making.

Administrative tools that operate without processing confidential manuscript content may be used where appropriately authorized.

Publisher-approved integrity screening systems may support editorial work under separately documented safeguards.

Access to confidential information through a technical service must comply with applicable privacy obligations, security requirements, and the journal's governance procedures.

The use of an authorized automated tool does not transfer responsibility for editorial decisions away from the designated editor.

Publisher-Approved Automated Integrity Screening

Sciences Force LLC may employ vetted tools to assist with legitimate publication integrity activities.

Examples include similarity screening, reference verification, duplicate-publication checks, image-integrity analysis, and identification of statistical or computational anomalies.

The use of such systems must be consistent with applicable privacy, confidentiality, intellectual property, and service-provider requirements.

Before confidential information is processed, the publisher must assess the system's purpose, data handling, access controls, retention practices, contractual safeguards, and relevant risks.

Automated screening results must be assessed by a qualified human editor or another appropriately authorized specialist.

A software alert is not a confirmed finding of research misconduct.

The publisher must not misrepresent automated screening as independent scientific peer review.

The use of an approved integrity tool does not authorize reviewers to use generative AI for confidential manuscript evaluation.

AI-Generated Peer Reviews and Manipulation of Editorial Processes

NOIS prohibits fabricated or improperly generated peer-review reports, false reviewer identities, and attempts to manipulate editorial decisions through automated systems.

Reviewers must not submit AI-generated evaluations as if they represent independently performed scholarly assessment.

Authors and third parties must not use AI to impersonate reviewers, create misleading professional identities, generate fraudulent editorial correspondence, or manipulate review recommendations.

Editors must remain alert to evidence of coordinated publication manipulation, including fabricated scientific content, misleading reference networks, purchased authorship, and automated mass production of unreliable manuscripts.

Such concerns must be investigated fairly and on the available evidence.

The use of automated tools alone is not proof of fraudulent activity.

AI Content Detectors and Editorial Assessment

NOIS does not regard automated AI-content detection scores as reliable proof that a manuscript was produced using generative AI.

Such scores must not be used as the sole basis for rejection, disciplinary action, or allegations of research misconduct.

AI detectors may produce false positives and false negatives, including for manuscripts written by non-native speakers of English or subsequently edited with ordinary language tools.

Where a legitimate concern arises, the editor should assess the manuscript's scientific content, supporting evidence, disclosure statements, methodological consistency, references, and relevant circumstances.

Authors may be asked for reasonable clarification when there is specific evidence of material undisclosed assistance or unreliable content.

The journal must not demand disclosure of private accounts, complete unrelated AI conversation histories, or confidential personal information without a legitimate and proportionate basis.

Editorial decisions must remain fair, evidence-based, and consistent with the journal's published policies.

Verification of AI-Assisted Research Outputs

Authors must validate material AI-assisted scientific outputs through appropriate independent procedures.

Verification should be proportionate to the type of research and may include:

- Mathematical proof checking.

- Independent numerical calculations.

- Code inspection and testing.

- Comparison with verified benchmark implementations.

- Statistical reanalysis.

- Validation of data transformations.

- Inspection of original images and processing records.

- Repetition of selected experiments.

- Evaluation of model uncertainty and robustness.

- Confirmation of source references and bibliographic information.

Authors should retain relevant evidence of verification and provide it when reasonably requested for editorial assessment.

The presence of an AI Use Statement does not establish scientific validity.

Likewise, publication of a manuscript does not imply that NOIS has independently reproduced every result unless a documented verification procedure has actually been performed.

Records, Traceability, and Research Reproducibility

Where AI materially affects scientific findings, authors must maintain records appropriate to the research.

Such records may include model information, relevant prompts, computational settings, data versions, generated outputs, processing scripts, verification results, and material human interventions.

Not every incidental interaction with an AI tool needs to be preserved.

Records should focus on information necessary to assess the reliability and provenance of the work.

Retention must be consistent with applicable institutional, legal, contractual, and data-protection obligations.

Where records contain sensitive or restricted information, secure storage and controlled access must be used.

AI-assisted computational methods must comply with the journal's Research Data, Code and Reproducibility Policy.

Reproducibility requirements must not be interpreted as authorization to disclose private or unlawfully obtained materials.

Submission Requirements and Editorial Declarations

Authors must provide an accurate declaration of relevant AI use when submitting a manuscript.

The corresponding author must ensure that the declaration reflects the activities of all contributors.

The disclosure must be updated where relevant AI assistance is introduced during revision, production, or preparation of a replacement manuscript.

A material change to the use of AI during scientific analysis requires an appropriate amendment to the Methods section and supporting documentation.

Where no AI use requiring disclosure occurred, the journal does not require a separate declaration of non-use.

The editorial office may request additional clarification if a declaration does not permit adequate assessment of the reported assistance.

Failure to provide an adequate declaration may result in a request for revision or further inquiry.

Publication depends on the scientific and ethical acceptability of the complete work, not merely the presence of a declaration.

Post-Publication Disclosure and Correction

Authors who identify a material omission or inaccuracy in an AI Use Statement after publication must notify the editorial office.

The journal will assess whether the omission affects scientific interpretation, methodological reproducibility, intellectual property rights, confidentiality, or the reliability of the reported results.

An incomplete disclosure does not automatically establish that the research is invalid.

Where the underlying findings remain reliable, an appropriate correction or disclosure update may be sufficient.

Where serious concerns remain unresolved, the journal may consider an Expression of Concern.

Where the findings are fundamentally unreliable or serious publication irregularities warrant action, a Retraction may be appropriate in accordance with the journal's publication ethics procedures.

Any action must be supported by evidence, evaluated proportionately, and linked transparently to the published record.

Allegations, Investigations, and Sanctions

Suspected misuse of artificial intelligence is assessed under the journal's Publication Ethics Statement and Complaints, Appeals and Whistleblowing Policy.

Relevant concerns include fabrication or falsification of research material, deliberate nondisclosure, intellectual property infringement, unauthorized disclosure of confidential information, fraudulent reviews, and systematic manipulation of publication processes.

The responsible editor may seek an explanation, supporting research records, independent expert assessment, or institutional clarification.

Authors and other affected individuals must normally be given an appropriate opportunity to respond.

Possible outcomes include clarification, correction, rejection, additional verification, expression of concern, retraction, or other proportionate action.

The journal must distinguish honest errors, incomplete disclosure, justified restrictions, and deliberate misconduct.

Serious or repeated misconduct may be referred to a relevant institution, funder, or competent authority when justified.

An AI-detection score alone is not sufficient evidence for imposing sanctions.

Accessibility and Fair Treatment

NOIS recognizes that AI-assisted technologies can support researchers with different language backgrounds, disabilities, and accessibility needs.

Responsible use of AI for translation, communication accessibility, or language clarity is not treated as scientific misconduct.

Applicable disclosure and verification requirements must nevertheless be respected.

Researchers must not be penalized merely for using permitted assistance or for writing in a style that resembles automatically generated text.

The journal assesses scientific validity, transparent methodology, originality, ethical compliance, and the accuracy of reported findings.

Editorial decisions must not be based on unsupported assumptions about a researcher's language background, nationality, institutional affiliation, or use of accessibility tools.

Policy Governance and Review

The Editor-in-Chief and Sciences Force LLC are responsible for implementing this policy within their respective editorial and publishing responsibilities.

The journal will review the policy as AI technologies, professional guidance, scientific practices, and applicable legal requirements evolve.

Substantive changes must be communicated through the official journal website and reflected consistently in related author, reviewer, editorial, and publication ethics policies.

The journal will not claim that particular AI screening, privacy protections, or independent verification procedures are operational unless they have actually been implemented.

Its central commitment is to encourage scientifically valuable research while ensuring that AI remains subject to transparent reporting, appropriate safeguards, and accountable human judgment.