(7 Apr 2025) Springer Nature has introduced a new AI-based tool to detect irrelevant references in submitted manuscripts, expanding its suite of in-house research integrity technologies.
The tool is being deployed by the publisher’s Research Integrity Group (RIG) across nearly all Springer Nature journals and books. It automatically analyzes the relevance of cited references in submissions and flags manuscripts with a high volume of irrelevant citations for further manual review by RIG. If confirmed, the submission may be withdrawn.
The reference checker has undergone multiple testing and validation phases to ensure accuracy across disciplines. Its development remains ongoing, and all use of the tool is subject to human oversight in accordance with Springer Nature’s AI Principles.
This launch follows the rollout of two earlier AI tools – Geppetto, designed to detect AI-generated fake content, and SnappShot, which identifies problematic images – both introduced within the past year. Together, these tools are part of the publisher’s broader strategy to safeguard research quality.
Reference analysis is a crucial method for detecting fraudulent submissions, particularly as AI-generated papers become increasingly sophisticated. The publisher maintains a zero-tolerance stance on fake research and continues to expand its efforts through technology, a dedicated 50-person integrity unit, and training programs for staff and editors.
Click here to read the original press release.
