This page states BMU-wide publishing principles. It does not replace journal-specific author guidelines, peer-review descriptions, fee schedules or discipline-specific ethical requirements. Operational claims apply only where implemented by the relevant journal or publisher system.
Human responsibility
AI systems are tools, not accountable authors, editors or reviewers. Humans remain responsible for the accuracy, originality, legality, ethical compliance, citations, data handling and final decisions associated with work produced or modified with AI assistance.
Disclosure
Material AI use should be disclosed when it contributes substantively to content generation, analysis, image creation, coding, interpretation or other scholarly output. Routine spelling, grammar or formatting assistance may be treated differently where it does not materially create intellectual content.
Each journal may specify where and how disclosure should appear, but the requirement should be clear enough for authors to understand before submission.
Confidential manuscripts
Unpublished manuscripts, reviewer reports, editorial deliberations, integrity allegations, identifiable participant information and other confidential material must not be uploaded to public or unauthorised generative-AI systems.
Use of a secure AI tool in confidential editorial work requires appropriate authorisation and safeguards concerning privacy, data retention, intellectual property and contractual obligations.
AI in peer review and editorial work
AI may assist limited administrative or analytical tasks where authorised, but it should not replace human scholarly judgment or act as the sole decision-maker for acceptance, rejection, complaints, misconduct findings or post-publication action.
Editors and reviewers remain responsible for verifying any AI-assisted analysis or language and for preserving confidentiality.
AI-generated or altered images
Where AI is used to generate or materially modify research-related images, figures or other evidential content, provenance and the nature of the alteration should be disclosed. Image enhancement must not create, remove or obscure features in ways that misrepresent the underlying evidence.
References and factual claims
Authors are responsible for verifying references, quotations, factual statements, calculations and code produced with AI. Fabricated citations or unsupported statements remain the responsibility of the human authors even when generated by a tool.
Participant and personal data
Sensitive personal data, identifiable participant information and protected health or research data should not be supplied to an AI system without a lawful and ethical basis and appropriate safeguards.
Policy evolution
AI technologies and associated standards change rapidly. BMU and its journals may revise this framework as tools, legal obligations and scholarly norms develop. Public wording should be updated when actual practice or approved tools change.
Policy basis and interpretation
Adapted principally from JHWCR Policy 10 and related confidentiality provisions. The BMU version generalises journal-specific material into a publisher framework and removes details that should remain title-specific. Where a journal publishes a more specific requirement that is consistent with these principles, the journal-specific rule governs that journal's submissions.
Questions about this framework may be sent to info@bmuventures.com.