SAICE Journal AI Policy

The South African Institution of Civil Engineering (SAICE) is committed to the responsible, ethical, and transparent use of artificial intelligence (AI) and AI-assisted technologies in its journal. This policy provides guidance to authors and reviewers on the appropriate use of AI tools.

1. Responsible use of AI and AI-assisted tools

AI and AI-assisted technologies may be used as supportive tools in the research, writing, reviewing and publication process. However, their use must remain ethical, transparent, and subject to meaningful human oversight at all times. Authors remain fully accountable for all content produced or evaluated with the assistance of AI tools, and peer reviewers remain responsible for the integrity, independence, and fairness of their evaluations and recommendations.

2. Guidelines for authors

Authorship

Authorship implies intellectual responsibility, accountability, and the ability to take responsibility for the integrity of the work. AI tools and systems cannot fulfil these requirements and must therefore not be listed as authors or co-authors.

Authors are responsible for:

• Producing original work that has not been published previously
• Ensuring that the work does not infringe on third-party rights
• Complying with the SAICE Journal author guidelines and this AI policy
• Taking full responsibility for all content in their manuscript, including any material generated or refined using AI tools

Use of AI in the writing process

Authors may use AI and AI-assisted technologies to improve clarity, grammar, language quality, or structure. Such tools may also be used for administrative or technical support (e.g. summarising notes or formatting text).

However, AI tools must not replace core scholarly activities, such as developing original ideas or hypotheses; interpreting data or results; drawing conclusions or recommendations; or exercising professional judgement.

Meaningful human oversight and editorial control must be maintained throughout the research and writing process. All AI-assisted content must be carefully reviewed, verified, and edited by the authors, who retain full responsibility for the accuracy, originality, and integrity of the work.

Disclosure

Authors must disclose any use of AI tools in the preparation of their manuscript at the point of submission. This disclosure must include the name(s) of the AI tool(s) used, and a brief description of how and for what purpose the tool(s) were used.

This declaration will be published alongside the article to ensure transparency.

The following table provides guidance on acceptable and unacceptable uses of AI tools during the preparation of manuscripts submitted to the SAICE Journal. These examples are not exhaustive but illustrate common use cases.

Use case	Allowed?	Disclosure required?	Notes
Generating substantive manuscript text using AI (e.g. drafting sections, literature summaries, or interpretations)	Not permitted	n/a
Creating images using generative AI tools (e.g. AI-generated figures, diagrams, or illustrations)	Not permitted	n/a
Using AI tools to modify images that contain research data (e.g. graphs, microscopy images, or experimental outputs)	Not permitted	n/a
Conception and design of study (e.g. summarising research trends or evaluating study protocols)	Allowed with disclosure	Yes
Editing, translating, or summarising author-written text to improve clarity or readability	Allowed with disclosure	Yes	AI tools may assist with editing but must not generate substantial scholarly content
Editing non-data images (e.g. conceptual illustrations, graphical abstracts, schematic diagrams)	Allowed with disclosure	Yes	Edits must not alter the scientific meaning of figures
Grammar checking and copyediting tools	Allowed	No
Literature searching or identifying relevant publications	Allowed with disclosure	Yes	AI tools can hallucinate citations; authors must verify sources
Similarity or plagiarism checking tools	Allowed	No	If AI is used to rewrite text based on similarity results, disclosure is required
Generating reference lists or bibliographies	Allowed with disclosure	Yes	Authors must verify all references
Programming or code generation (e.g. Python or MATLAB analysis code)	Allowed with disclosure	Yes	Authors must verify correctness of all code and outputs

Plagiarism and research integrity

Manuscripts submitted as original work must not contain AI-generated content presented as the author’s own intellectual output. The submission of AI-generated text, images, data, or analyses as original scholarly work constitutes a form of plagiarism and is inconsistent with SAICE’s ethical standards.

Authors are encouraged to use AI tools only as aids in refining their work and must apply critical thinking and professional judgement when reviewing and revising any AI-assisted output.

Compliance

Non-compliance with this policy may result in rejection, correction, retraction, or other editorial action.

3. Guidelines for reviewers

Considering the use of AI by authors

Reviewers must take into account any declared use of AI tools by authors when evaluating a manuscript.

• The use of AI has been appropriately disclosed
• AI tools were used in a supportive manner rather than generating original scholarly contributions
• The manuscript demonstrates clear human intellectual input and professional judgement
• The use of AI has not compromised the originality, accuracy, or integrity of the work

If reviewers suspect undisclosed AI-generated content, they should raise these concerns in confidential comments to the moderator or Editor-in-Chief.

Use of AI in the review process

Peer review is a human-centred scholarly activity. AI tools must not replace the reviewer’s expertise or judgement.

• AI must not generate substantive review content
• AI must not assess novelty or scientific merit
• AI must not make accept/amend/reject recommendations
• AI must not draft reviewer reports without human oversight

The final review report must reflect the reviewer’s independent assessment.

Accountability and ethical responsibility

Reviewers remain fully responsible for the content of their review reports and must ensure evaluations are fair, unbiased, constructive, and based on their expertise.

4. Policy development and updates

SAICE recognises the dynamic nature of AI technology and reserves the right to review and update this policy to reflect technological developments, legal requirements, and best practice.