Artificial Intelligence Disclosure in Academic Nursing: A Framework for Editorial Policy and Practice

Integrating artificial intelligence (AI) tools into academic publishing is transforming how manuscripts are prepared, reviewed, and evaluated. From grammar assistance to content generation, AI offers robust solutions to enhance writing clarity, streamline workflows, and analyze complex data. These advancements not only bring new ethical challenges, particularly regarding transparency and disclosure [1], but also promise to improve the efficiency and quality of scholarly publishing significantly. Nurse editors are uniquely positioned to address these challenges in nursing research, ensuring that AI use in manuscripts aligns with the integrity, accountability, and credibility underpinning nursing science and scholarly publishing. One of the primary ethical concerns is the lack of clarity in journal guidelines regarding using and disclosing AI tools. While many nursing journals address AI use and disclosure in their journal guidelines (often adapted from publisher recommendations), they frequently lack the robustness necessary to prevent unethical AI use. A recent study used ChatGPT to analyze journal policy and discovered several themes, including the need for transparency, disclosure, and, most importantly, a need to update journal policies that remain ambiguous [2]. Additionally, many journals permit grammar assistance tools without disclosure, but the line becomes obfuscated as AI tools increasingly assist with rewriting, paraphrasing, and generating new content powered by AI integration. Similarly, AI use in research workflows, such as data analysis, raises questions about Institutional Review Board (IRB) or ethics committee approvals and whether authors accurately reported their processes. Undisclosed or unethical AI use risks undermining the trustworthiness of nursing scholarship and poses a significant threat to evidence-based practice and the nursing scientific record, laying bare the urgent need for vigilance in this area [3]. Today, nurse editors use tools to scan manuscripts for potential AI use, much the same way editors use tools to detect plagiarism. However, while plagiarism is a concern with AI-generated content, detecting AI use in manuscripts is not as straightforward as detecting plagiarism, making the use of these tools somewhat outdated and, in some cases, obsolete. For example, studies have shown that AI-detecting software continues to produce inaccurate results, even with 100% human-written content [4, 5]. These challenges signify the need for practical guidance instead of solely utilizing electronic detection tools. These challenges are not new concerns as the scholarly community continues to grapple with manuscript authorship and accountability. For example, in response to these issues, the International Committee of Medical Journal Editors (ICMJE) provided initial guidance for editors, which many nursing editors continue to rely on. According to ICMJE guidelines, authorship requires individuals to contribute significantly to the conception, design, execution, or revision of the manuscript while maintaining accountability for its content. AI-generated text does not meet these authorship criteria, as AI lacks intellectual ownership, responsibility, and the ability to approve final manuscript versions. ICMJE's position aligns with broader concerns about the potential for misattribution when AI-generated content is integrated without proper disclosure [6]. Nursing journals, many of which follow ICMJE guidelines, should ensure that AI use is transparently disclosed while preventing the improper attribution of AI-generated content as human-authored work. These emerging problems open the door for a framework to further support nursing editors as they evaluate submissions. A growing body of literature on this topic in various scientific disciplines further supports the need for a framework specific to nursing. For example, Hosseini et al. [7] recently published an article arguing that outright bans on AI-generated text are neither enforceable nor effective. The authors also advocate for transparency of use beyond a disclosure statement. In fact, they suggest that authors document the use of AI in their methods sections or introduction sections. They also suggest that authors provide in-text citations to track use across the scholarly community. However, the latter suggestion seems increasingly impossible when and/or if authors use AI to generate the bulk of their literary texts. These background and seminal works collectively guide this paper's suggestions in the nursing context. This paper presents a simplified decision-making framework for nurse editors to evaluate manuscripts where AI use is disclosed and where its use is suspected. By aligning the framework with ethical publishing principles, including guidance from the Committee on Publication Ethics (COPE), ICMJE, and contemporary literature, this work aims to provide actionable strategies for promoting transparency, ensuring compliance, and addressing the ethical implications of AI in academic nursing. Importantly, this paper emphasizes the pivotal role of nurse editors in developing robust editorial policies to address AI use, empowering them to navigate this evolving landscape while safeguarding the credibility of nursing research. As AI becomes integral to scholarly publishing, nurse editors are passive observers and active shapers of the ethical standards guiding its responsible use. Transparency and disclosure are central to the COPE guidelines, particularly with technologies like Large Language Models (LLMs). With the rapid rollout of LLMs like ChatGPT, COPE adopted the stance that AI tools cannot be listed as authors because they lack the capacity for accountability, decision-making, and ethical responsibility [8]. However, while this principle addresses authorship concerns, it does not sufficiently handle the broader ethical implications of AI use in writing, data analysis, or research workflows. The proposed framework in this paper expands on COPE's emphasis on transparency by requiring detailed disclosure of AI use, regardless of its role, and aligning these disclosures with journal policies. COPE's focus on integrity in research and reporting aligns with the framework's recommendations for explicit and accurate descriptions of AI use in research processes. While COPE emphasizes the importance of ethical research practices, it does not currently provide detailed guidance on using AI in research workflows, such as data analysis, statistical modeling, or qualitative coding procedures. Using AI-powered tools for qualitative analysis calls into question various privacy concerns for participants' intimate data. This lack of specificity creates a gap that this framework seeks to address by requiring authors to disclose AI use in research workflows and confirm oversight, such as Institutional Review Board (IRB) approval, where applicable. This raises ethical concerns, particularly regarding the accuracy, validity, and privacy of data processed by AI. Recent discussions in the academic literature discuss the risks of AI-generated biases, potential errors in data interpretation, and privacy vulnerabilities when using AI in human-subject research [7]. Ensuring that AI-driven methodologies are ethically validated and appropriately disclosed is critical to upholding the reliability of findings and preventing the introduction of biases, errors, copyright issues, and privacy issues into the nursing literature. Education and guidance, cornerstone principles of COPE, are reflected in the framework's emphasis on equipping editors, reviewers, and authors with the tools to navigate the ethical challenges of AI in publishing. While COPE encourages education and the development of proactive policies, it has not yet addressed the need for specific frameworks to manage the complexities of AI use in manuscripts. This framework responds to this gap by recommending targeted training for editors and reviewers to identify potential AI use and assess its ethical implications. Additionally, it calls for journals and publishers to provide clear policies and educational resources for authors on proper AI use and disclosure. By extending COPE's principles into actionable strategies, this framework ensures that AI is used responsibly and integrated into the publishing process in a way that upholds the values of transparency, accountability, and integrity. This proactive approach fills a critical void, offering nurse editors a structured pathway to address the ethical and practical issues surrounding AI in academic publishing. As it becomes more common, practical, and acceptable to disclose AI use in manuscripts, editors are alerted to examine the quality and reliability of the manuscript's findings and primary conclusions more closely. This framework recommends a stepwise approach when the author already knows and discloses AI use. First, the editor must confirm the scope of the disclosed use. If the authors disclosed their use, were they specific about how and why they used AI? Did they only use it for writing improvement, flow, or clarity? Transparency about contributions, including those made by AI, is a core principle of ethical publishing. COPE emphasizes that authors must disclose all significant contributions. Editors can request detailed statements describing the specific uses of AI in the work. After doing so, the editor must confirm that the reported use aligns with the journal and publisher's current written policies. Journal editors should work with their publishers and editorial board members to develop policies requiring full disclosure, not just how the tools were used for writing assistance but when and if they were used in data analysis. The next step is to address transparency. In aligning with COPE standards, disclosure is not enough. When disclosure is known, the journal editor might require a specific statement within the manuscript instead of just inside the editorial management system. This statement would improve transparency by clarifying that the author disclosed AI use within the work. With this transparency statement, the author should fully address potential limitations, bias, and mitigation strategies. Lastly, the editor should evaluate the ethical implications of AI use based on these disclosure statements. Some journal guidelines prohibit AI-generated images, while others only make disclosure provisions when AI is used for writing clarity and flow. Journal editors should ask if the researcher used AI in any research process, particularly data analysis. Suppose the author discloses AI use beyond just writing assistance. In that case, the editor should confirm that an ethics committee or IRB knows the researcher's use of AI in research processes. This author's experience indicates that many US-based IRB panels do not specifically ask researchers if their protocols include using AI for data analysis. I strongly suspect that if researchers did disclose this, the IRB would disapprove because the researcher would need to detail their privacy mitigation strategies, which will likely prove impossible. If the author cannot confirm that the IRB approved their use in research processes, the editor should consider a rejection based on ethical violations. The fact that these specifics are not addressed in journal policy leaves an open door for misconduct in research. When evaluating manuscripts that authors did not disclose AI use, editors should look for red flags first. These indicators include inconsistent writing style, overly polished language, or repetitive phrases characteristic of AI-generated text. Use AI detection tools cautiously, understanding their limitations and inaccuracies. Aligning with COPE recommendations, editors are encouraged to investigate concerns that might indicate ethical issues, such as undeclared AI use. Consider policy updates, including training editors and reviewers to recognize potential signs of undeclared AI use and handle these cases with due diligence. Avoid over-reliance on AI detection software and incorporate manual review processes. If undisclosed AI use is not suspected, the editor can proceed with the review process if the manuscript meets the quality guidelines and fits the journal's aim. In instances where the editor or editorial team suspects undisclosed AI use, the editor must investigate further. COPE encourages fairness and dialog when handling suspected ethical concerns, encouraging editors to engage with authors transparently. Reach out to authors for clarification about their writing or research processes and possible AI use. Maintain a non-punitive approach unless there is clear evidence of deliberate misconduct. Incorporating more transparency in the editorial policy may lead to more instances where the editor must investigate further and, as such, should have a procedure for these investigations. If the editor confirms undisclosed or unethical use of AI in submitted manuscripts, they should educate and support the authors following COPE recommendations. COPE emphasizes fostering ethical behavior and educating authors about publication ethics. Suppose AI use is confirmed but not disclosed. In that case, the editor should inform authors about the importance of transparency, provide resources on ethical AI use, and encourage revisions to ensure compliance with journal standards. The editor might also choose to educate and reject the work in collaboration with the editorial team, board, and/or publisher. A proactive approach to policy development is essential to handling AI's challenges in academic publishing. Transparency is a cornerstone of ethical publishing, as COPE emphasizes, and journals must require detailed disclosures of all AI tools used during manuscript preparation or research workflows. Policies should explicitly state that authors must disclose AI use for writing assistance, editing, or research purposes. The author's guidelines should include examples of acceptable AI use, such as grammar correction, and unacceptable use, such as generating entire text sections without disclosure. Providing these clear distinctions will promote trust and reduce ambiguity. This framework also recommends clarifying expectations for AI use in research. COPE's principles on ethical research call for accurate and transparent reporting of methodologies, which extends to AI-driven processes like data analysis or statistical modeling. Journals should require authors to disclose the use of AI tools in the methodology section and provide details on their validation. Additionally, IRB approval should be required for studies involving AI-driven research processes to ensure oversight and adherence to ethical standards. These policies will ensure that AI use does not compromise the credibility or integrity of research findings. To further support transparency, authors should be required to include a dedicated section on transparency and ethical considerations in their manuscripts (or other such naming conventions). This section should outline the use of AI tools, describe their roles, and address any potential limitations or biases the researcher encountered. The authors should also detail their privacy mitigation strategies to ensure human subject safety. Establishing procedures for handling suspected undeclared AI use is equally important. COPE encourages fairness and transparency in addressing ethical violations, so journals must develop clear protocols for investigating such cases. These protocols should include steps for author communication, resolution, and consequences for non-disclosure, ranging from mandatory revisions to withdrawal of the manuscript. The COPE website provides a few editor-reported cases of undisclosed AI use and shows how to handle these challenges. Additionally, educating editors and reviewers is essential to executing these policies effectively. Training sessions on identifying and addressing ethical concerns related to AI use can help editors and reviewers navigate these emerging challenges. Providing checklists that include AI-related considerations could further standardize the evaluation process. Finally, plagiarism and copyright policies should be revised to address AI use explicitly, ensuring that AI-generated content does not inadvertently plagiarize or violate intellectual property. Combining plagiarism detection tools with manual review processes will enhance the protection of originality and integrity. These recommendations, grounded in COPE's ethical principles, will ensure that journals are equipped to address the complexities of AI use in academic publishing. Table 1 provides an overview of the major points of this framework. Request detailed disclosure of AI tools used (e.g., writing assistance, data analysis). Verify adherence to journal policies on AI use (e.g., grammar vs. rewriting). Require explicit AI disclosure in submission forms, specifying tools and their purpose. Add a disclosure template or form to submission requirements. Ensure inclusion of a transparency and ethical considerations section in the manuscript. Verify acknowledgment of limitations or biases introduced by AI tools. Mandate a standardized disclosure section addressing AI use. Confirm findings and conclusions are not misrepresented due to AI use. For AI in research (e.g., data analysis), ensure IRB approval is explicitly stated. Provide clear examples of acceptable and unacceptable uses of AI in writing and research in journal guidelines. Look for signs of AI use, such as abrupt changes in writing style, overly polished language, or repetitive phrasing. Train editors and reviewers to recognize potential AI-generated content. Avoid over-reliance on AI detection software. Use AI detection tools cautiously; confirm suspicions through dialog with authors. Maintain a non-punitive approach unless there is clear evidence of deliberate misconduct. Develop protocols for contacting authors about suspected AI use, emphasizing education over punishment. Offer guidance and resources on ethical AI use. Encourage revisions to ensure compliance with journal standards. Include AI-related education materials in journal author guidelines. Offer workshops or webinars for authors and reviewers. In the spirit of disclosure, the author used Grammarly's new AI-integrated feature, “Increase the impact of your text,” during the writing and preparation of this text; as such, some parts of the paragraphs in this paper were reconstructed using AI-generated content. The author carefully reviewed each area of reconstructed content to ensure accuracy. The author did not use AI tools to cite resources found in this work. This framework was developed through the experience of this author's evaluation process of submitted manuscripts. Lastly, in the spirit of transparency, the author is a nursing editor but reports no further conflicts of interest. This framework offers a first step for editors seeking guidance in the uncertain environment of rising AI use in nursing research. Implementation of the framework should involve several journal stakeholders, including the editors, editorial board members, society (if applicable), and publishers. It is also important to note that as AI evolves, so should this framework and journal policies and guidelines. Transparency, disclosure, and upholding ethical standards are key to ensuring academic and research integrity when evaluating manuscripts that could be published in the nursing literature. The risk of not taking these simple implementation steps and policy and training suggestions is immense. Nursing science is at risk with the rise of AI in scholarly outlets. At the time of this writing, the full impacts on nursing science and evidence-based practice are unknown as AI continues to evolve and entrench itself in scholarly and academic spaces and workflows. Editors, reviewers, researchers, and authors should remain vigilant, open-minded, and flexible as we navigate these exciting yet uncertain times. The author declares no conflicts of interest.