image: Dr Milos Cuculovic, head of technology innovation at MDPI
Credit: MDPI
MDPI, the open access publisher, has today announced the full deployment of its in-house AI-powered research integrity system, Ethicality. The tool is now being used to automatically screen all manuscripts submitted to the publisher.
Its roll-out marks a significant step forward in MDPI’s commitment to safeguarding the scientific record, improving efficiency, and supporting editorial decision making from initial submission through to the publication decision. Ethicality now screens in the region of 2,000 submissions from authors worldwide per day.
Ethicality is an integrated, end-to-end integrity layer within the editorial workflow, continually monitoring manuscripts rather than acting as a one-time screening tool. Its development has been shaped by extensive testing and iterative training, allowing MDPI to refine detection capabilities based on real-world submission patterns. Its deployment comes at a time of significant interest regarding the use of AI in scholarly publishing.
“What is becoming clear is that traditional, manual processes are no longer sufficient in peer review. The industry needs to shift from reactive approaches, resolving issues after publication, to proactive systems that support editors earlier in the workflow. AI, when used responsibly, acts as a set of guardrails rather than a substitute for human judgment. The future lies in combining automation with strong editorial oversight to ensure consistency, transparency, and trust at scale,” says Dr Milos Cuculovic, head of technology innovation at MDPI.
“The publishing industry is undergoing a fundamental shift driven by scale and technology. Submission volumes continue to grow, while expectations around speed, transparency, and quality are increasing. At the same time, generative AI is creating both opportunities and risks—from improved workflows to challenges such as synthetic content, manipulated data, and questionable authorship practices,” said Dr Cuculovic.
AI-Driven Integrity Screening at Scale
Ethicality analyses each submission according to core components—such as title, abstract, author metadata, main text, and references—before conducting a comprehensive integrity assessment. Peer review reports are also analysed as part of the process. It screens for a wide range of potential issues, including:
- Paper mill activity and fabricated submissions;
- AI-generated or manipulated text;
- Citation manipulation and irregular referencing patterns;
- Fake reference detection;
- Author identity concerns and authorship anomalies;
- Suspicious peer review patterns and AI-generated text in peer review reports.
Separately, MDPI uses third party software such as Proofig to manage image manipulation and iThenticate to scan all submitted manuscripts for text duplication and potential plagiarism.
Supporting Human Editors Through Automation
Ethicality was created by MDPI to serve as a support system rather than a replacement for editorial decision-making. A core principle of Ethicality is its human-in-the-loop design. While the system provides automated analysis and risk signals, all flagged cases are reviewed by experienced human editors or research integrity professionals before any action is taken. This approach ensures that automation enhances, rather than replaces, editorial oversight.
Dr Enric Sayas, product owner of Ethicality, said: “We are in a technological race. As generative AI makes it easier to produce sophisticated plagiarism and high-quality fake papers, traditional detection methods are no longer sufficient. The only viable response is to deploy equally advanced tools—using large language models and specialized AI systems to detect manipulated images, inconsistent data, and AI-generated content. Without such safeguards, the volume of fraudulent submissions risks overwhelming peer review and undermining the credibility of scholarly publishing.”
By automating time-intensive technical checks, the system allows editors to focus on scientific assessment and decision quality.
“The primary value of AI is its ability to handle time-consuming aspects of manuscript processing, allowing editors and reviewers to focus on high-level scientific evaluation. Tasks such as reference validation, formatting checks, and basic technical triage are essential but repetitive. AI can perform an initial screening, flagging problematic cases for editorial review, ensuring that human expertise is applied where it matters most—scientific decision-making,” said Dr Sayas.