A framework for sustainable, human-centered integration of AI is proposed in which AI is restricted to technical verification and efficiency, while judgments on scientific merit, ethics, and paradigm-shifting research are reserved for appropriately valued human experts.
Abstract
The integration of Large Language Models (LLMs) and Artificial Intelligence (AI) into scientific publishing is accelerating, driven by systemic crises in peer review and academic economic pressures. This manuscript, presented as a critical commentary and policy analysis, provides a three-part examination: a) AI as a co-authoring tool, balancing its democratizing potential against risks like citation fabrication, b) AI's proficiency in technical review versus its inability to assess novelty, and c) the risk of amplified bias in AI-driven editorial decisions. In the context of peer review, LLMs are being increasingly used for tasks such as preliminary technical verification and language editing. However, their adoption raises critical concerns. Recent evidence confirms that citation hallucination remains a persistent threat, although emerging mitigation strategies such as Retrieval-Augmented Generation (RAG) and verified knowledge integration are being developed to address this challenge. Furthermore, 2025 surveys indicate that over 50% of researchers utilize AI during peer review, often violating existing policies. This shift occurs within an exploitative model that relies on unpaid labor while charging substantial Article Processing Charges (APCs). To address these challenges, this paper proposes a framework for sustainable, human-centered integration of AI. A hybrid model is proposed in which AI is restricted to technical verification and efficiency, while judgments on scientific merit, ethics, and paradigm-shifting research are reserved for appropriately valued human experts. The economic feasibility of compensating reviewers, the allocation of verification responsibilities within editorial workflows, and the need for clear legal accountability frameworks are examined. Maintaining scientific credibility requires both the ethical integration of AI and a fundamental reform of the economic structures governing research communication.
Abstract The use of artificial intelligence (AI) tools is transforming how research is planned, conducted, written, and reviewed, offering opportunities but also imposing limitations, particularly regarding scientific integrity. This editorial note addresses the uses and limitations of AI tools in the preparation of scientific articles. Key possibilities include support for writing and editing, translation, journal selection, literature review, citation and reference management, and the production of summaries and visualizations. Conversely, the limitations relate to a lack of creativity, the perpetuation of biases and inequalities, the risk of hallucinations (false information), issues of originality and intellectual property, and authorial responsibility. It advocates for the adoption of best practices in the use of AI tools in the preparation of scientific articles. Such use must be guided by transparency, human supervision, and accountability; it must be disclosed to editors at the time of article submission and, when relevant, mentioned in the article itself. Although AI tools are useful assistants in the preparation of scientific articles, they should not replace the researcher’s critical thinking and scientific knowledge.
A. Brandão, Leila Posenato Garcia, Luisa Veras de Sandes-Guimarães· Revista Brasileira de Saúde...· 0 citations
Medical publishing has undergone one of the most profound transformations in its history. From traditional manuscript handling to electronic submission systems, digital peer review, plagiarism detection software, and open-access publishing, each technological advance has reshaped scholarly communication by improving the efficiency, accessibility, and dissemination of scientific knowledge. The emergence of generative artificial intelligence (AI), AI-powered evidence-synthesis platforms, and increasingly autonomous AI assistants marks the next stage in the evolution of scholarly publishing.¹ Unlike earlier innovations that primarily improved publishing processes, these technologies increasingly assist with and, in some cases, undertake tasks that have traditionally depended on human intellectual judgement, including literature discovery, scientific writing, data analysis, peer review, editorial decision support, and knowledge synthesis. Consequently, the challenge confronting medical journals has shifted from adopting new technologies to governing their responsible use in ways that foster innovation while safeguarding scientific integrity, editorial accountability, and public trust. Artificial intelligence is reshaping medical publishing through three complementary technologies. Generative AI models, including OpenAI's ChatGPT (GPT-4o and GPT-5), Google's Gemini, Anthropic's Claude, and Meta's Llama, generate original text, code, tables, figures, and multimedia content. AI-powered search and evidence-synthesis platforms, such as Perplexity AI, Elicit, Consensus, Semantic Scholar AI, and Google AI Overviews, assist with literature retrieval, evidence synthesis, and knowledge discovery. Increasingly autonomous AI assistants, including Microsoft Copilot, ChatGPT Agent, and Gemini Workspace, have evolved from conversational tools into workflow assistants capable of performing multi-step tasks such as literature review, manuscript drafting, statistical coding, journal selection, and preparation of responses to peer reviewers. Although frequently grouped together under the umbrella of AI, these technologies present fundamentally different editorial challenges.²
Kamran Khalid· Journal of Fatima Jinnah Med...· 0 citations
The integration of artificial intelligence (AI) into scholarly peer review represents a fundamental transformation of academic publishing’s quality control mechanisms. This report critically examined the ethical considerations, institutional practices, and emerging technologies associated with AI-assisted peer review. Drawing on recent policy documents from major publishing organizations, empirical research on AI implementation, and critical scholarship on algorithmic bias, this analysis revealed significant tensions between efficiency gains and integrity preservation. While AI tools have demonstrated potential for addressing reviewer burnout and publication delays, their deployment raises critical concerns regarding confidentiality breaches, accountability gaps, algorithmic bias, and the erosion of expert judgment. Major organizations (e.g., International Committee of Medical Journal Editors) and leading publishers such as Elsevier and Taylor & Francis have emphasized disclosure where AI is used, human accountability, and strict confidentiality controls—often prohibiting uploading unpublished manuscripts into generative AI tools. However, empirical evidence has suggested nontrivial, and potentially growing, undisclosed large language model (LLM)-assisted text in peer review in some conference contexts. This report concluded that AI should serve as an augmentative rather than substitutive technology in peer review, with robust governance frameworks, transparent disclosure mechanisms, and continuous evaluation of equity implications essential for responsible implementation.
Aras Bozkurt· International Review of Rese...· 0 citations
BACKGROUND
There has been an increase in the availability and use of artificial intelligence (AI) across many professional domains, and there are traces of AI in nearly every manuscript published using modern technology. It is becoming increasingly difficult for peer reviewers, editorial teams, and journal readers to identify the degree to which authors have used AI in the development of their manuscripts.
METHODS
Our research group developed an AI scoring rubric to provide authors with an opportunity to self-disclose their use of AI.
RESULTS
The Transparency and Reporting of Artificial Intelligence Contribution for Evaluating Submissions (TRACES) instrument provides a score from 0 to 40 across 3 domains: mechanics, writing, and illustrations. Higher scores indicate increased use of AI by the authors when writing or preparing a manuscript for submission.
CONCLUSION
Authors in any field should self-report a TRACES score when submitting their manuscript. Journals may benefit from requiring authors to include a TRACES score when submitting a manuscript for peer review. While higher TRACES scores indicate greater use of AI, there is no specific cutoff provided to determine manuscript acceptance or rejection.
M. B. Vyas, Lori M Rhudy, Suzanne Weckman et al.· Journal of Neuroscience Nurs...· 0 citations
Artificial intelligence (AI) is changing electronic word-of-mouth by participating directly in the production, revision, circulation, evaluation, and governance of consumer information. This systematic literature review synthesizes 71 peer-reviewed studies on how AI functions within electronic word-of-mouth (eWOM), how consumers respond to those roles, and when those effects vary. Using SPAR-4-SLR and PRISMA 2020 procedures, the review draws on database, supplemental, and citation-based searches. It identifies five recurring AI roles -- Creator, Co-Creator, Curator, Intermediary, and Regulator -- and shows that their effects depend largely on perceptions of authenticity, credibility, usefulness, humanness, transparency, and manipulation risk. These effects are context dependent, varying across products, tasks, message formats, consumer orientations toward AI, levels of AI involvement, and platform environments. The review also argues that AI reshapes not only how eWOM is presented, but also what consumers are likely to contribute afterward. Building on these findings, the review proposes an AI-enabled lifecycle of Creation, Transformation, Transmission, Evaluation, and Governance and develops an AI-eWOM fit perspective. A TCCM-organized research agenda identifies priorities for future research.
A. Joyal· Journal of business and mana...· 0 citations