General Principles
Artificial intelligence and AI-assisted technologies may provide useful support in research and scholarly communication. Their use, however, does not reduce the responsibility of human researchers, authors, reviewers or editors for the work they produce and the decisions they make.
AI tools should be used in a manner consistent with research integrity, transparency, confidentiality, intellectual property rights and applicable ethical requirements.
Individuals using AI-assisted technologies remain responsible for verifying the accuracy, originality, appropriateness and integrity of any material generated or modified with their assistance.
Permitted AI-Assisted Use
Authors may use AI-assisted technologies for appropriate supporting purposes provided that their use does not compromise the integrity, originality or reliability of the work.
Such use may include assistance with language, grammar, readability, organization, translation, coding or other research-related tasks where the authors retain appropriate human oversight and independently verify the resulting material.
Authors should not rely on AI-generated output as a substitute for appropriate scholarly judgement, methodological expertise, verification of evidence or critical interpretation of research findings.
Disclosure of AI Use
Material use of generative AI or AI-assisted technologies in preparing or conducting work submitted for publication should be disclosed transparently where the use is relevant to understanding how the manuscript, analysis or research was produced.
Where disclosure is required, authors should identify the tool or technology used and describe its purpose sufficiently clearly to allow editors and readers to understand the nature of its role.
Where an AI system forms part of the research methodology itself, its use should be described in the appropriate methodological section with sufficient information to support transparency and, where relevant, reproducibility.
Routine tools that provide basic spelling, grammar, formatting or similar conventional assistance generally do not require the same level of disclosure as substantive generative AI use, unless their role materially affects the scholarly content.
AI in Research
Where artificial intelligence is used as part of the research process, authors should ensure that its use is scientifically appropriate and adequately described.
Relevant information may include the system or model used, its role in the study, important settings or procedures and the manner in which outputs were evaluated or validated.
Authors should consider potential limitations, errors, bias and uncertainty associated with AI systems and should not present AI-generated results as independently verified evidence unless appropriate validation has been performed.
Where the use of AI involves personal, confidential, sensitive or otherwise protected data, authors are responsible for ensuring that appropriate ethical, legal and institutional requirements are satisfied.
AI-Generated and AI-Modified Images
The use of artificial intelligence to create, alter or enhance images, figures or other visual material must not result in misleading representation of research findings.
AI tools must not be used to fabricate, selectively alter, remove, add or obscure features of research images or data in a manner that misrepresents the underlying evidence.
Where AI-generated or substantially AI-modified visual material is legitimately part of the research or scholarly work, its nature and method of creation should be disclosed where relevant.
Routine technical adjustments should remain consistent with accepted disciplinary standards and must not change the interpretation of the underlying data.
Accuracy, Sources and References
Authors must independently verify factual claims and references suggested or generated by AI systems. AI-generated citations should never be assumed to be genuine or accurate without verification against reliable scholarly sources.
Fabricated references, incorrect bibliographic information, invented quotations, unsupported claims or nonexistent data remain the responsibility of the authors if included in a submitted manuscript.
Authors should consult and cite the original scholarly sources that support their work rather than treating an AI-generated response as a substitute for the underlying evidence.
Confidentiality, Privacy and Data Protection
Authors should consider carefully whether information entered into an external AI system may be retained, processed, reused or disclosed by the service provider.
Confidential manuscripts, unpublished research, personal data, proprietary information, sensitive research data or other protected material should not be submitted to an AI service where doing so would breach confidentiality, privacy, consent, contractual obligations or applicable legal or ethical requirements.
Responsibility for determining whether use of a particular AI service is appropriate rests with the individual using the service.
Use of AI by Reviewers
Manuscripts provided for peer review are confidential documents. Reviewers must protect manuscript content and associated information from unauthorized disclosure.
Reviewers must not upload a confidential manuscript, substantial portions of a manuscript, unpublished data or other protected submission material to an external generative AI system where confidentiality or data protection cannot be assured.
Reviewers remain personally responsible for the content, reasoning and recommendations contained in their review. AI-generated output should not replace the reviewer's independent scholarly assessment.
Any permitted use of AI assistance must remain consistent with the journal's confidentiality, peer-review and disclosure requirements.
Use of AI by Editors
Editors and editorial staff are responsible for protecting the confidentiality of submitted manuscripts and information obtained through the editorial process.
Confidential manuscript content or unpublished material should not be entered into external generative AI systems where doing so could expose the material to unauthorized processing, retention or disclosure.
Editorial decisions must remain under meaningful human responsibility. Automated systems may provide appropriate administrative or analytical support, but they should not independently make final decisions concerning acceptance or rejection of manuscripts.
Editors remain responsible for evaluating the evidence, reviewer reports and relevant editorial considerations when making publication decisions.
AI Detection and Screening
Science Insights may use appropriate screening or editorial tools to identify potential concerns relating to submitted material. Results produced by automated detection systems should be interpreted cautiously and in context.
An automated indication that content may have been generated or modified by AI does not, by itself, establish misconduct or inappropriate AI use. Where a concern is significant, additional assessment and clarification may be required.
Editorial action should be based on relevant evidence and the circumstances of the case rather than solely on the output of an automated detection tool.
Misuse of Artificial Intelligence
AI technologies must not be used to fabricate or falsify data, images, references, reviewer identities, peer-review reports, ethical documentation or other information associated with research or publication.
AI-assisted plagiarism, deliberate misrepresentation of research, manipulation of the peer-review process or other deceptive practices may constitute research or publication misconduct.
Suspected serious misuse may be considered under the Research Integrity and Misconduct Policy and may result in appropriate editorial action.
Evolving Technologies
Artificial intelligence technologies and their applications in research and publishing continue to develop rapidly. Science Insights may revise this policy as technologies, scholarly practices, ethical expectations and relevant standards evolve.
Authors, reviewers and editors are expected to exercise appropriate professional judgement when encountering uses of AI that are not specifically addressed by this policy and to prioritize transparency, confidentiality, accountability and the integrity of the scholarly record.