A US-developed AI has identified over 1,000 fake scientific journals by analysing website patterns: Heres how its helping to clean up academic publishing

Researchers at the University of Colorado Boulder have developed a US-based AI tool that identified over 1,000 potentially fake scientific journals by analysing website features such as editorial board credibility, citation patterns, and publishing frequency. Trained on data from the Directory of Open Access Journals, the AI flagged over 1,400 suspicious journals out of 15,200 assessed. Human experts later confirmed the majority were predatory. The system is intended to assist in safeguarding academic publishing from fraudulent practices.