Trust Signals (Datavid)

DatavidBuilt by: companyStatus: BetaYear: 2026

Overview

Trust Signals is a research-integrity screening system from Datavid that checks a submitted manuscript’s authors, references, affiliations and disclosures against research metadata, combining the results into a single, explainable Trust Score from 0 to 100. It uses more than 30 trust markers grouped into four areas: author credibility and verification, reference quality and integrity, institutional affiliations, and transparency and compliance. Its data comes from ORCID, ROR, Crossref and OpenAlex.

The checks cover problems such as identity mismatches, retracted or off-topic references, funding and clinical trial disclosure issues, and affiliation anomalies. The system also compares an author’s previous work with the topic of the submitted paper, and it analyses citation and collaboration networks to identify patterns of risk across groups of authors.

Datavid says the score is meant to show editors which manuscripts need closer review before peer review begins, with the evidence behind each flag, and that publishers can tune the weights and thresholds to their editorial policies or risk tolerance. In 2026 Datavid opened a limited early access programme in which a small group of publishers are evaluating the scoring.

Notes and quotes

  • Datavid says Trust Signals automates checks that take “hours if not days”.
  • The Trust Score is shown in four bands: 85–100, low risk, for standard review; 70–84, moderate risk, for enhanced editorial review; 50–69, elevated risk, for deeper investigation; and below 50, high risk, where comprehensive analysis is needed.
  • Elaine Ellerton, a former scientific editor who writes for Datavid: “A risk score should not act as a verdict.”

Sources

This entry was written with the help of AI, from the sources listed. Please check the sources before relying on it, and let us know if anything needs correcting.