AllegationVault
A free, press-grade archive where journalists and researchers can search a database of public-record allegations against public figures, with linked primary sources and a confidence score for each claim — built to prevent the rumors that swirl around cases like Busfield's.
journalists and researchers checking allegations before publishing
- Searchable index keyed to public-record sources (police reports, court filings, motions) with provenance links on every claim
- Confidence tags: alleged, charged, adjudicated, overturned — auto-applied from filing metadata
- Embeddable citation widgets journalists can paste into articles to show their sourcing
- Browser extension that flags social posts making claims without a primary source attached
Coverage like USA Today's Busfield story both depends on and contradicts claims sourced only to testimony; a public-record-first tool serves both reporters writing the story and readers trying to evaluate it.
CourtListener (Free Law Project) handles ~7.9M queries and 24M API calls per 10 days explicitly serving journalists, lawyers, academics — proves sustained demand for public-record search, though no evidence of dedicated 'allegation' databases with confidence scoring.CourtListener / Free Law Project ↗
CourtListener covers case law but not allegations-as-entities with confidence scores; adjacent projects (Little Sis, ICIJ, Influence Explorer) cover influence/finance, leaving a narrow gap for a press-grade allegation archive — narrower than the initial 7 estimate.CourtListener / Free Law Project ↗
Positioned as free; CourtListener sustains itself on donations and premium memberships but reports no advertising or transactional revenue — strong evidence that free public-record journalism tools monetize weakly.CourtListener / Free Law Project ↗
Public records are permanent, journalism verification needs are evergreen, and the Busfield-style news cycle creates recurring relevance — solid long-term utility but no evidence of accelerating demand.
CourtListener proves indexing hundreds of jurisdictions is feasible open-source, but adding structured 'allegation' objects plus editorial confidence scores requires significant human curation, lowering feasibility from the initial 7.