Journal Of Colloid And Interface Science AI Policy: ChatGPT and Generative AI Disclosure Rules for JCIS Authors

JCIS authors should check four policy areas under Elsevier's current AI framework before submission. The point is to map each AI use to a manuscript surface before upload, not to paste one generic disclosure sentence at the end of drafting.

Rule 1: Disclose every AI tool used in manuscript preparation

Authors should document substantive generative-AI use with the tool name, version or access date, and how it was used. Elsevier's JCIS guide asks authors to add a declaration section before the references when AI tools were used in manuscript preparation, and to describe research-method AI use in Methods when it affects data, code, analysis, or image interpretation. Examples that require disclosure at JCIS:

• using ChatGPT, Claude, Gemini, or similar to draft, polish, translate, or substantially edit manuscript text passing through JCIS editorial review

• using AI to generate limitations, data-availability language, graphical-abstract captions, or JCIS-specific response-to-reviewers text tied to Elsevier's framework

• using AI to translate manuscript text into English from another language, with the source language and translation chain documented

• using AI for citation discovery, literature screening, or summarizing prior JCIS work, because AI-generated references can be incomplete or fabricated

• using AI-assisted code, curve-fitting scripts, image-analysis workflows, or adsorption-model calculations that affect reported results

Examples that do NOT require AI disclosure:

• At JCIS, using grammar/spell checkers (Word) for line-level edits, when used without generative AI features for new manuscript content

• For JCIS submissions, using reference managers (Zotero, EndNote) for citation formatting against Elsevier's style guide

• For Journal of Colloid and Interface Science (Elsevier) statistical analysis, using established statistical software (R, Stata, SPSS) where the algorithm is the established tool documented in JCIS's methodological norm, not a generative AI

Rule 2: AI cannot be an author

No AI tool can be listed as an author of a JCIS paper, particularly for colloid and interface science research with quantified surface-property characterization and mechanistic interpretation-class submissions. Under Elsevier's policy: authorship requires the ability to take responsibility for the content, agree to be accountable for accuracy, and to consent to publication. AI tools cannot do any of these in JCIS's editorial framework. This rule is consistent across all Elsevier-published journals and applied at JCIS's desk-screen.

Rule 3: Generative AI and figures, images, and artwork need a separate check

Elsevier's journal guidance does not permit generative AI or AI-assisted tools to create or alter images in submitted manuscripts, including moving, obscuring, removing, enhancing, or introducing image features. The stated exception is when AI use is part of the research design or methods; then the use must be described reproducibly in Methods with the model or tool, version and extension numbers, manufacturer, and the role of the tool. For JCIS, treat microscopy images, adsorption plots, particle-size distributions, contact-angle photographs, graphical abstracts, and composite figures as higher-risk than prose because the figure package often carries the scientific claim.

Rule 4: Disclose AI use in peer review participation

Reviewer AI-use rules are publisher-specific and can change quickly. Reviewers must follow the journal's confidentiality and AI-use policy; authors should not assume that reviewer-side AI rules are identical across journals in the same portfolio.

Acceptable disclosure language for JCIS submissions:

"For our colloid and interface science research with quantified surface-property characterization and mechanistic interpretation-focused manuscript at JCIS, we used ChatGPT-4o (OpenAI, version dated October 2024) to polish English-language phrasing in the Introduction and Discussion sections. We did not use generative AI for data analysis, figure generation, or substantive manuscript content. All authors reviewed and edited the AI-assisted text and take responsibility for the final manuscript."

Or, for AI-assisted code:

"For this JCIS submission addressing colloid and interface science research with quantified surface-property characterization and mechanistic interpretation, initial Python code for the Bayesian regression analysis was drafted with Claude 3.5 Sonnet (Anthropic, version dated December 2024). All code was reviewed, modified, and validated by the authors before use; the final version is available at [repository URL]. Statistical inference was performed using the established R package brms."

What does NOT pass JCIS's desk-screen:

• For JCIS addressing colloid and interface science research with quantified surface-property characterization and mechanistic interpretation: "AI tools were used in manuscript preparation." Too vague for Elsevier editorial review of JCIS submissions; the JCIS editorial team needs the specific tool name, version, and specific use case

• "We acknowledge AI assistance in the Acknowledgments." (Do not rely on this location unless the current journal policy explicitly allows it.)

• "ChatGPT helped write this paper." (Insufficient detail on use case)

• No disclosure when AI was used (publication-ethics violation)

For JCIS-targeted manuscripts, the patterns below are common AI-policy risk areas to check against the publisher's current guidance before submission because unresolved AI-use questions can escalate from an editorial query into rejection, correction, retraction, or investigation when they affect figures, code, authorship, or the scientific record.

AI disclosure missing despite obvious AI-assisted mechanism language. Substantive AI-assisted drafting without a required declaration can trigger an editorial query. In our JCIS reviews, this usually appears as polished but generic explanations of interfacial mechanism, adsorption kinetics, or particle-stability behavior that do not match the specificity of the data package. Check whether your manuscript reads as AI-assisted

AI declaration placed in the wrong manuscript location. Elsevier's JCIS guide asks for a declaration section before the references for manuscript-preparation AI use, while AI involvement in research methods belongs in Methods. A cover-letter-only note or a vague acknowledgement can fail because it does not tell reviewers whether AI touched the evidence they are evaluating.

Check whether your AI disclosure is in the right section

Generic declaration language without tool name, version, and affected surface. "AI tools were used" is too thin when the paper contains AI-shaped benchmark language, fitting code, graphical-abstract wording, or literature-comparison logic. The JCIS version of the problem is specific: the declaration should say whether the tool touched the figure, Methods workflow, code, caption, literature screen, or only prose. Check whether your AI disclosure has the required specificity

AI-assisted figures treated like ordinary layout work. Elsevier's figure policy is stricter than a formatting note. If AI created or altered an image, graphical abstract, or evidential figure, the manuscript needs a policy check before upload; if AI was part of the research method, the Methods section needs reproducible tool detail.

For Journal of Colloid and Interface Science authors, an incomplete AI declaration can lead to a submission query, returned files, a Methods clarification request, rejection, correction, retraction, investigation, or deeper publication-ethics review if AI affected data, images, code, authorship, or the scientific record. Elsevier says manuscripts may be checked with screening tools, and its image guidance allows image-forensics review for suspected irregularities.

The higher-risk cases are incomplete declarations around original figures, AI-assisted analysis code, fabricated or unchecked citations, and language that makes the contribution look broader than the actual colloid or interface evidence. Unauthorized authorship changes can lead to rejection or retraction under the JCIS guide; undisclosed AI use should be treated with the same seriousness when it affects accountability or image integrity.

In our pre-submission review work for JCIS and peer colloid/interface venues, the AI-policy compliance gap most consistent across the cohort is generic declaration language without tool-version and manuscript-surface specificity. The field-specific version matters because AI often touches zeta-potential interpretation, adsorption-model wording, graphical abstracts, figure captions, microscopy-processing descriptions, or code-assisted fitting steps. Those are not cosmetic surfaces. They affect how editors and reviewers judge mechanism, reproducibility, and novelty.

Three JCIS-specific failure patterns recur:

JCIS adsorption-model declaration mismatch

The manuscript reports Langmuir, Freundlich, kinetic, or regeneration results, but the AI-assisted wording turns a limited fit into a broad mechanism claim. The disclosure problem is not only language use; it is that the reviewer cannot tell whether AI helped reshape a mechanistic interpretation.

JCIS microscopy or graphical-abstract disclosure gap

Authors sometimes disclose AI as "writing assistance" even when a figure label, schematic, composite image, or graphical abstract was AI-assisted. For JCIS, that can affect how the editor reads particle morphology, interfacial structure, or the claimed evidence chain.

JCIS code-assisted fitting hidden under prose editing

AI may have drafted Python, MATLAB, or R code for curve fitting, image measurement, or property prediction, but the declaration only says the text was edited. If the code influenced a plotted result, the Methods section needs the tool role and author validation path.

JCIS literature-screening benchmark drift

A JCIS paper can look stronger or weaker depending on which adsorption, emulsion, membrane, or nanoparticle papers are used as comparators. If AI helped screen or summarize the literature, the declaration should make that role visible instead of presenting the benchmark table as fully manual.

We check JCIS AI declarations against the parts of the manuscript that create the strongest reviewer reliance. If AI helped draft a benchmark table comparing adsorption capacity, the declaration should say whether AI structured the comparison or only polished wording. If AI helped write code for an isotherm fit, the Methods section should separate code drafting from author validation. If AI touched a graphical abstract or microscopy panel, the figure-policy check happens before the manuscript is uploaded.

For concrete JCIS evidence surfaces, think about adsorption and interface articles such as DOI 10.1016/j.jcis.2024.10.176, DOI 10.1016/j.jcis.2025.138409, or DOI 10.1016/j.jcis.2025.138965. Those DOI examples are not AI-use claims; they are Crossref-returned examples of the kind of JCIS article structure where an undisclosed AI role in figures, code, or literature comparisons would materially affect editorial trust.

The practical Manusights screen is therefore not "did you use ChatGPT?" It is: did AI touch a figure, method, data transformation, code file, literature-search step, benchmark claim, or interfacial-mechanism paragraph that a reviewer needs to evaluate? If yes, the declaration needs to be specific enough that a JCIS editor can see the tool, date or version, role, author verification, and manuscript surface without asking a follow-up question.

Journal of Colloid and Interface Science (Elsevier) follows the publisher's public AI policy, but authors should verify the current journal page before submission because AI-use rules, disclosure locations, and image guidance continue to change.

Elsevier's AI policy framework continues to evolve as 2026 brings new ICMJE recommendations, COPE guidance refinements, and journal-specific clarifications. JCIS authors targeting colloid and interface science research with quantified surface-property characterization and mechanistic interpretation submissions should track three signals throughout 2026:

Quarterly policy updates from Elsevier. The publisher's public AI policy guidance is updated over time. JCIS authors who pre-register their disclosure language at submission time tend to face fewer revisions during the 2026 transition period than authors who write boilerplate disclosures.

Field-specific clarifications for colloid and interface science. Different research domains see different AI use patterns. JCIS authors should treat AI involvement in adsorption models, interfacial-energy interpretation, microscopy workflows, and code-supported characterization as higher risk than line editing. Authors who err on the side of precise disclosure avoid the publication-ethics gray zone.

Elsevier sister-journal consistency. Authors often compare JCIS with 5 sister journals or nearby Elsevier titles: Colloids and Surfaces A, Applied Surface Science, Chemical Engineering Journal, Journal of Molecular Liquids, and Materials Chemistry and Physics. The publisher policy umbrella is shared, but the evidence surface changes by journal, so copy-pasting the same AI declaration across targets is risky.

Reviewer disclosure norms. As Elsevier extends AI-disclosure rules to peer reviewers, the response rate from JCIS reviewers may shift. Authors should expect that JCIS reviewers' use of AI tools is now also disclosed and factored into editorial decisions.

• Manusights internal preview corpus (2025 cohort)