Artificial Intelligence and Generative Technologies

Open Christian Press permits responsible, critically supervised use of artificial intelligence and automated tools where such use strengthens legitimate scholarly or production work without displacing human accountability, violating confidentiality or misrepresenting the origin and reliability of content.

Technology does not become responsible merely because it is efficient. Its use must be judged by purpose, transparency, accuracy, proportionality, legality and its effect on persons and the scholarly record.

Human accountability

AI systems cannot be authors, reviewers, editors or rights holders because they cannot approve a manuscript, disclose conflicts, consent to publication or accept responsibility for error or misconduct. Every named human contributor remains accountable for the content submitted or published under that person's authority.

Acceptable support uses

Subject to disclosure and verification requirements, AI tools may assist with:

·       Grammar, spelling, readability and language refinement

·       Translation support reviewed by a competent human

·       Coding, data cleaning or analytical exploration under human supervision

·       Transcription or summarisation of authorised material

·       Organisation of literature or notes

·       Generation of preliminary outlines, questions or alternative formulations

·       Production tasks such as metadata checking, accessibility support or format conversion

These uses do not transfer intellectual responsibility to the tool.

Prohibited or unacceptable uses

Authors must not submit fabricated references, invented quotations, false data, synthetic participant accounts, manipulated evidence or unverified claims generated by AI. AI output must not be presented as empirical findings or primary evidence unless the research explicitly studies such output and reports the method transparently.

AI must not be used to imitate a real author's work deceptively, create unauthorised images of identifiable persons, conceal plagiarism, manufacture reviewer reports or overwhelm the Press with mass-produced submissions.

Disclosure by authors

Material use of AI should be described in the manuscript or declaration, identifying the tool, purpose and extent of use where relevant to evaluation or reproducibility. Routine spelling correction or basic formatting assistance need not normally be disclosed unless the journal or book programme requires it.

Disclosure does not excuse error. Authors must verify citations, calculations, quotations, code, images and interpretations independently.

Confidentiality for editors and reviewers

Editors and reviewers must not upload unpublished manuscripts, identifiable author information, reviewer reports, personal data or restricted files to public AI services without explicit authority and appropriate safeguards. Many such services may retain, process or reuse submitted content in ways incompatible with peer-review confidentiality.

Private or institutionally approved systems may be used only where the Press has assessed the data, contractual and security implications.

AI-assisted peer review

AI may support administrative tasks, language clarification or structured checking, but it must not replace the reviewer's independent reading and scholarly judgement. Any material AI assistance in producing a review should be disclosed to the editor.

The reviewer remains responsible for the report and must ensure that automated suggestions do not introduce false claims, bias or confidential disclosure.

AI-generated images and media

Synthetic images, diagrams, audio or video must be identified where their artificial origin could affect interpretation. Authors are responsible for lawful training-source concerns, permissions, representation of persons, accuracy of scientific or historical detail and avoidance of deceptive manipulation.

Detection tools

AI-detection software may be used as a limited screening signal but not as conclusive proof of misconduct. Such tools can produce false positives and may disadvantage multilingual authors. Decisions require contextual human assessment and an opportunity for explanation where appropriate.

Editorial response

Undisclosed or irresponsible use may lead to clarification, revision, rejection, correction, retraction or referral under the publication ethics policy. The response should reflect the seriousness, intent, effect and reliability of the evidence.

The governing principle is simple: tools may assist human scholarship, but they may not replace truthfulness, judgement, consent or responsibility.