How We Verify
Our four-tier source verification system
How We Verify
Every article published on this platform undergoes a rigorous automated credibility assessment. Our verification system is designed to be transparent, reproducible, and independent.
The Four-Tier Source System
Tier 1 — Wire Services: Reuters, Associated Press, AFP, and official government records. These carry the highest weight due to strict editorial standards and minimal retraction rates.
Tier 2 — Major Outlets: BBC, The New York Times, Financial Times, Bloomberg, The Guardian, and other internationally recognised publications with established editorial processes.
Tier 3 — Specialty Sources: Regional outlets, trade publications, and subject-matter-specific media such as CoinDesk, Politico, and MarketWatch.
Tier 4 — Unverified: Social media posts, blog content, and sources without established editorial oversight.
Scoring Methodology
Each article receives a composite score (0-100) based on four factors: source corroboration (30%), source tier quality (25%), individual claim verification rate (30%), and source recency (15%). The final score determines the credibility label displayed to readers.
Claim-Level Verification
Articles are decomposed into individual factual claims. Each claim is independently assessed as Confirmed, Likely, Unverified, Disputed, or False, based on how many and which tier sources support it.
Community Voting
In addition to AI-generated credibility scores, readers can vote on whether they consider a story true, unsure, or false. This community signal operates independently of the algorithmic assessment.