AI-assisted editorial screening
Purpose and scope. Informatics in Education (INFEDU) may use journal-controlled software and approved artificial-intelligence or large-language-model services, including approved API services, to support the preliminary editorial and technical assessment of submitted manuscripts.
AI-assisted screening may process the whole de-identified manuscript or selected manuscript sections, tables, references, figures, and extracted metadata where necessary. It may support:
- scope and manuscript-type assessment;
- submission-completeness and reporting checks;
- methods, results, and claims-consistency checks;
- reference, citation, and semantic literature checks;
- numerical and table-consistency checks;
- research-integrity and ethics risk indicators;
- reviewer-expertise matching and editorial routing; and
- preparation of confidential internal editorial summaries.
Human editorial control. AI-generated observations are advisory only. INFEDU does not accept, reject, rank, or otherwise decide the outcome of a manuscript solely through automated processing. All editorial assessments, correspondence, reviewer assignments, and decisions are made or approved by qualified human editors. Editors are responsible for checking material AI-generated observations against the manuscript and other relevant evidence.
Confidentiality and data minimisation. Submitted manuscripts remain confidential editorial materials. Where practicable, author names, affiliations, contact details, document metadata, acknowledgements, and other unnecessary identifiers are removed before external AI processing. The separate title page is not normally included in AI-assisted manuscript assessment.
INFEDU seeks to avoid transmitting unnecessary personal data, directly identifying participant information, special-category data, confidential third-party information, or restricted research materials. Authors should therefore include only information necessary for scholarly evaluation in the anonymised review manuscript.
Approved providers and permitted use. Manuscript content may be processed only through INFEDU-approved accounts, software, locally operated models, or contracted API providers. Personal editor accounts and unauthorised public AI systems must not be used for confidential manuscript processing.
Approved external providers act under their applicable contractual, confidentiality, data-protection, security, and retention terms. INFEDU does not intentionally use submitted manuscript content to train general-purpose artificial-intelligence models.
Where the OpenAI API is used, OpenAI states that API inputs and outputs are not used to train its models by default. Provider-side abuse-monitoring or temporary processing retention may nevertheless apply unless enhanced contractual retention controls have been enabled.
Providers, models, prompts, and screening procedures may be changed, suspended, or replaced as editorial and technical requirements evolve, subject to applicable institutional, contractual, ethical, and legal requirements.
Retention and editorial records. INFEDU minimises external storage of manuscript content and uses available provider controls to restrict unnecessary retention. Internal screening summaries, editor verifications, and related actions may be retained as confidential parts of the journal's editorial record in accordance with operational, contractual, legal, and research-integrity requirements.
Accuracy and limitations. AI systems may produce incomplete, inaccurate, inconsistent, or unsupported observations. AI-assisted screening does not guarantee that plagiarism, fabricated data, citation problems, methodological errors, security issues, or other integrity concerns will be detected. It is not a substitute for qualified editorial assessment or external peer review.
Internal AI-assisted reports are editorial working materials. They do not constitute peer-review reports, do not create an entitlement to acceptance or external review, and are not normally supplied to authors or reviewers.
Information security and limitation. INFEDU and contracted service providers apply reasonable technical and organisational safeguards appropriate to the editorial workflow. Nevertheless, no internet transmission, software platform, API, cloud service, or information system can be guaranteed to be completely secure, continuously available, or free from unauthorised interference.
To the fullest extent permitted by applicable law, INFEDU and its editors make no warranty that third-party technology services will be uninterrupted, error-free, or immune from malicious acts, unauthorised access, transmission failures, or other events beyond reasonable control. Responsibility and liability for any incident are determined by applicable law and the relevant contractual arrangements. Nothing in this notice excludes any right, duty, or liability that cannot lawfully be excluded.
Author acknowledgement. Submission to INFEDU constitutes acknowledgement that the manuscript may undergo the AI-assisted processing described in this notice as part of the journal's editorial administration, screening, quality-control, and research-integrity procedures.
Authors whose manuscripts contain information subject to special legal, contractual, national-security, commercial-confidentiality, or data-transfer restrictions should contact the Editorial Office before submission. INFEDU may be unable to process a manuscript if such restrictions are incompatible with the journal's editorial workflow.
Questions concerning this policy may be sent to editorial@infedu.lt .