Wikipedia 'AI-Generated' File Caption Drift
Tracking how quickly AI-generated captions on Wikipedia drift into inaccuracy or hallucination.
This subject is not publishing figures yet — its collector is still being built.
No figures yet
This subject has been created but its first collection run hasn't produced data yet.
About this data
This page monitors AI-generated image captions on Wikipedia, measuring how often they contain hallucinated terms and how long they persist before being corrected. The data reveals the reliability of automated descriptions in one of the web’s most trusted knowledge sources, highlighting the speed at which AI-generated errors or outdated information can spread. By tracking caption age and accuracy decay, we provide insights into the lifespan of AI-generated content in a high-stakes environment.
Why this isn't published anywhere else
No existing source tracks the specific 'lifespan or accuracy decay' of AI-generated file captions on Wikipedia as a dedicated, recurring metric or dashboard.
Uniqueness score 1.00 — assessed against live web search results when this subject was created.