How we supervise AI enrichment before it reaches the reader.
Editorial Lead, NarrativeRadar
M. Reyes leads the editorial function at NarrativeRadar. The role oversees the rules that AI enrichment must follow, reviews narrative comparisons before they reach readers, and signs off on corrections. The aim is the same across every event: separate observable facts from competing interpretations and conditional scenarios, and never present any of them as the truth.
NarrativeRadar generates most of its content with language models. Editorial oversight is the set of rules and reviews that make that content publishable: which event types we cover, which claims are admissible, which tone is banned, and which signals force an abstention.
Every event runs through three automated checks before it reaches the homepage — source quality, strict separation of facts/narratives/scenarios, and editorial-language filter. When a user reports an error or two events are merged, the editorial lead reviews and authorises the correction or merge, which is logged publicly at /corrections.
What we do not do: we never label a narrative as true, false, or disinformation; we never recommend buying, selling, or allocating capital; we never present a scenario as a certainty. The full rules are at /editorial-policy.