Public claims organised into intelligence categories.
Categories are not just navigation labels. They are memory partitions for tracking claims, contradictions and narrative evolution within specific public domains.
AI
Model capabilities, benchmarks, safety narratives, compute claims and infrastructure economics.
Politics
Campaign claims, institutional messaging, policy narratives and political contradiction memory.
Economics
Inflation, growth, labour, productivity and forecasting narratives.
Health
Medical evidence, public-health claims, pharmaceutical narratives and intervention evidence.
Science
Research claims, replication, evidence quality and scientific interpretation.
Climate
Climate projections, emissions claims, sustainability evidence and environmental narratives.
From AI claim testing to broader public-claims intelligence.
The original EvidTrace methodology was developed through frontier AI-company claim analysis. That work remains the proving ground and scoring foundation, now generalised into a wider institutional intelligence system.
Capture
Identify public claims and preserve them as canonical records.
Evidence
Attach evidence summaries, provenance and verification context.
Contradict
Track when later claims or evidence weaken earlier narratives.
Remember
Build claimant reliability and confidence trajectories over time.
Reason
Use relationship graphs to analyse narrative propagation.
EvidTrace is not trying to become another stream of news. It is trying to preserve continuity in an environment where stories move faster than public memory.
The long-term aim is to help people see what was claimed, who claimed it, what evidence existed, what later contradicted it and how the surrounding narrative evolved.