YOUR GO-TO INTELLIGENCE RESOURCE FOR THE SCHOLARLY PUBLISHING INDUSTRY

MDPI has introduced full deployment of its in-house AI-powered research integrity system, Ethicality, which now screens all manuscripts submitted to the publisher. The system processes approximately 2,000 submissions per day from authors worldwide.

Ethicality functions as an integrated integrity layer within the editorial workflow, continuously monitoring manuscripts rather than serving as a one-time screening tool. Its development involved extensive testing and iterative training, refining detection capabilities based on real-world submission patterns. The launch coincides with growing interest in the use of artificial intelligence in scholarly publishing.

MDPI representatives noted that traditional manual processes are increasingly insufficient for peer review. They emphasized that the publishing industry is shifting from reactive approaches, which address issues after publication, to proactive systems that support editors earlier in the workflow. According to MDPI’s head of technology innovation, AI can serve as a safeguard when combined with editorial oversight, ensuring consistency and transparency at scale. He also observed that rising submission volumes, coupled with expectations for speed and quality, are reshaping publishing practices, while generative AI introduces both opportunities and risks.

Ethicality analyzes multiple components of each submission—including title, abstract, author metadata, main text, references, and peer review reports—to conduct integrity assessments. It screens for issues such as:

• Paper mill activity and fabricated submissions

• AI-generated or manipulated text

• Citation manipulation and irregular referencing patterns

• Fake references

• Author identity concerns and authorship anomalies

• Suspicious peer review patterns, including AI-generated content

In addition, MDPI continues to use third-party tools such as Proofig for image manipulation detection and iThenticate for plagiarism checks.

Ethicality was designed to support, not replace, editorial decision-making. The system follows a human-in-the-loop model, with flagged cases reviewed by editors or research integrity professionals before action is taken. MDPI’s product owner for Ethicality explained that generative AI has made it easier to produce sophisticated plagiarism and fabricated papers, requiring equally advanced detection tools. He highlighted that without such safeguards, fraudulent submissions could overwhelm peer review and undermine credibility.

By automating technical checks, Ethicality enables editors to focus on scientific assessment and decision quality. MDPI representatives stated that AI’s primary value lies in handling repetitive tasks such as reference validation and formatting checks, allowing human expertise to be applied to scientific evaluation.