OpenAlex is considering adding a novelty score to its platform based on a winning approach from the Metascience Novelty Indicators Challenge. The Jülich team's method uses large language models to analyze papers and their citations to assess novelty at publication time, providing scores, confidence intervals, and written justifications. The score aims to fill a gap where citation counts measure usefulness but not newness, though the team emphasizes it should support rather than replace human judgment.