Skip to main content

​
Three-Layer Verification System

ViewFT’s verification system combines three independent layers to assess content accuracy. No single layer has authority alone — the weighted consensus determines the outcome.

​
Layer 1: AI Fact-Check (30% weight)

What it does: Automated cross-referencing of claims against trusted source databases, news archives, and fact-checking organizations. Strengths: Fast (seconds), scalable, consistent, no human bias. Limitations: Can miss nuance, satire, opinion framing, and emerging stories with limited source material. This is why it’s only 30% — AI provides a first signal, not the final word.

​
Layer 2: The Senate (30% weight)

What it is: An elected editorial board of high-ViewCred community members who review disputed content. How Senators are elected:
  • Candidates must have Authority-level ViewCred (600+)
  • Election held quarterly via ViewCred-weighted community vote
  • Senate consists of 7-11 members (initially smaller, growing with the community)
  • Senators serve 6-month terms with a maximum of 2 consecutive terms
How review works:
  • A panel of 3 randomly selected Senators reviews each disputed piece
  • Senators provide a written assessment with reasoning
  • Simple majority determines the Senate’s position
  • Senators who consistently vote against the eventual consensus lose credibility

​
Layer 3: Community Vote (40% weight)

What it is: Open vote by all eligible community members, weighted by ViewCred Score. Why it has the highest weight: This aligns with ViewFT’s decentralization principles. The community — not algorithms, not a small board — has the final say. How voting works:
  • Voting window opens for 24-72 hours after AI and Senate layers complete
  • Each vote is weighted by the voter’s ViewCred Score
  • Minimum quorum required for vote to be valid
  • Results are transparent and published

​
Weighted Outcome

The final determination combines all three layers: 
Final Score = (AI Score × 0.30) + (Senate Score × 0.30) + (Community Score × 0.40)
  • Score > 50%: Content marked as misleading, slashing applied
  • Score ≤ 50%: Content cleared, flaggers may lose ViewCred for false flagging
  • Score 45-55% (ambiguous): Flagged for additional review, no immediate slashing

​
Design Philosophy

Why three layers instead of one?
  • AI alone would miss nuance and context
  • Senate alone could become a centralized censorship body
  • Community alone could be gamed by coordinated groups
  • Together: each layer checks the others’ blind spots