How Research and Science Today navigates the intersection of AI innovation and scholarly integrity — and what it means for our authors, reviewers, and readers.
How Research and Science Today navigates the intersection of AI innovation and scholarly integrity — and what it means for our authors, reviewers, and readers.
Artificial intelligence is not a future concern for scholarly publishing — it is a present reality. AI tools are already embedded in the workflows of authors, reviewers, editors, and publishers. They assist with language editing, detect plagiarism, identify suitable reviewers, screen for image manipulation, and flag submissions suspected of originating from paper mills. At the same time, the same technology creates new risks: fabricated content, undisclosed AI-generated manuscripts, compromised peer review, and the erosion of accountability that underpins the scholarly record.
This article outlines how Research and Science Today approaches these challenges — not with blanket prohibitions, but with a principled framework grounded in the standards of the Committee on Publication Ethics (COPE) and the practices of leading international publishers.
Our Guiding Principles
Transparency over prohibition
We do not ban AI. We recognize that generative AI tools offer legitimate benefits — particularly for non-native English speakers, for managing large datasets, and for accelerating routine aspects of manuscript preparation. What we require is transparency. Authors must disclose how AI was used. Reviewers must not delegate their judgment to machines. And we, as editors, must be honest about the AI-assisted tools we use in our own editorial processes.
Human accountability is non-negotiable
Every manuscript published in this journal has human authors who have approved the final version, who stand behind its claims, and who can be held accountable for its integrity. Every peer review is written by a qualified human expert who has read the manuscript and applied their professional judgment. AI may assist in both processes, but it does not replace the human responsibility at their core.
Integrity screening is proactive, not reactive
We do not wait for problems to surface after publication. Our editorial workflow now incorporates the Ethical Shield v2 platform, which performs systematic checks at the point of submission — including author identity verification through the OpenAlex scholarly database, overlap detection for potential duplicate submissions, bibliographic coherence analysis, numerical consistency screening, and review quality assessment. These checks are designed to identify problems before they enter the peer review process, not after they have contaminated the scholarly record.
The Threat Landscape
Paper mills and fabricated research
Paper mills — organizations that produce fraudulent manuscripts for sale — represent one of the most serious threats to scholarly integrity today. In 2023, a major publisher was forced to retract over 8,000 articles traced to paper mill activity. AI-powered detection tools have since identified hundreds of thousands of additional published papers bearing textual signatures associated with paper mill production. At Research and Science Today, our screening tools analyze submission patterns, author networks, reference validity, and textual indicators to flag potentially fraudulent submissions before they reach reviewers.
AI-generated manuscripts
The concern is not that AI is used to improve a manuscript — it is that AI might be used to create one from scratch, without genuine research behind it. A manuscript generated entirely by AI, submitted by an author who did not conduct the described research, is not a scholarly contribution — it is fabrication. Our editorial screening includes analysis of whether submitted manuscripts contain patterns characteristic of AI-generated text, and we reserve the right to request evidence of underlying research, raw data, and methodological documentation at any stage of the review process.
Compromised peer review
AI-generated peer reviews that mimic the language of expert evaluation without providing genuine critical assessment represent a growing concern across the industry. Our review quality assessment module evaluates every review for specificity, anchoring in manuscript content, and consistency with the reviewer’s recommendation. Reviews that appear generic, formulaic, or disconnected from the manuscript are flagged and may be replaced with assessments from additional reviewers.
AI as an Ally in Editorial Integrity
It is important to acknowledge that AI is not only a source of risk — it is also an increasingly powerful tool for safeguarding integrity. AI-powered systems can detect image manipulation that is invisible to the human eye, identify citation networks associated with fraudulent activity, and screen thousands of manuscripts for textual patterns indicative of paper mill production. These capabilities did not exist a decade ago. They are now essential to maintaining the quality of the scholarly record at scale.
At Research and Science Today, we use AI-assisted tools in our editorial workflow for the following purposes:
- Plagiarism screening. All submissions are checked using similarity detection tools.
- Author verification. Author identities and publication histories are validated against the OpenAlex scholarly database.
- Review quality assessment. Reviewer reports are analyzed for specificity and anchoring in manuscript content.
- Submission integrity checks. Manuscripts are screened for patterns associated with duplicate submissions, self-plagiarism, numerical inconsistencies, and bibliographic incoherence.
These tools support editorial decision-making. They do not replace it. Every recommendation generated by our systems is subject to human editorial review and judgment.
What This Means for Our Community
For authors
Use AI tools if they help you do better work — but be transparent about it. Disclose what you used and how. Verify everything the AI produces. Take full responsibility for your manuscript. Understand that undisclosed AI assistance, if detected, will be treated as a breach of publication ethics.
For reviewers
Your expertise is irreplaceable. AI cannot evaluate whether a research question is important, whether a methodology is appropriate for a specific context, or whether conclusions are proportionate to the evidence. We ask you to protect the confidentiality of the manuscripts you review, to provide specific and constructive feedback, and to disclose any AI assistance used in preparing your review report.
For readers
Every article published in Research and Science Today has undergone rigorous human peer review, supported by AI-assisted integrity screening. We are committed to ensuring that the work we publish is original, methodologically sound, ethically conducted, and transparently reported. Our editorial processes are aligned with COPE guidelines and are subject to continuous improvement as the technological landscape evolves.
Looking Ahead
The relationship between AI and scholarly publishing will continue to evolve. New tools will emerge. New risks will surface. Policies that are appropriate today may need revision tomorrow. What will not change is our commitment to the foundational principles of publication ethics: honesty, transparency, accountability, and the primacy of human judgment in evaluating scholarly work.
We invite our authors, reviewers, and readers to engage with these issues openly. If you have questions about our AI policies, concerns about a submitted or published manuscript, or suggestions for how we can improve our processes, we welcome your communication. The integrity of the scholarly record is a shared responsibility — and it is one we take seriously.
— — —
This article is informed by the COPE Position Statement on Authorship and AI Tools, the COPE Focus on Artificial Intelligence and data from the 2025 International Congress on Peer Review and Scientific Publication. Research and Science Today operates in full compliance with COPE principles.