Why a Trust Score?
As AI agents proliferate across the internet, every platform faces the same question: should I trust this agent? API keys prove authorization but say nothing about an agent's behavior history, compliance posture, or operational integrity. The DRD Score is a 0-100 composite metric that answers this question quantitatively. It distills an agent's entire governance history into a single number that platforms can use for access decisions, rate limiting, and trust-weighted interactions.
Score Components
The DRD Score is calculated from five weighted components: Compliance History (30%) — the agent's track record of policy adherence over time, including violation frequency, severity, and recency. Identity Verification (20%) — the strength of the agent's identity credentials, from basic registration to full W3C VC with DID resolution. Policy Coverage (20%) — how comprehensive the agent's governance policies are relative to its operational scope. Behavioral Consistency (15%) — statistical analysis of the agent's action patterns, flagging anomalous behavior. Transparency (15%) — the completeness of the agent's audit trail and willingness to share compliance proofs.
Scoring Algorithm
Each component produces a sub-score from 0-100. The final DRD Score is a weighted average with decay functions applied to older data. Recent compliance matters more than historical compliance — an agent that was perfect for six months but had three violations last week scores lower than one with steady, moderate compliance. The decay function uses an exponential curve with a half-life of 30 days. Scores are recalculated every hour for active agents and daily for inactive ones. The algorithm is deterministic — the same inputs always produce the same score.
Trust Tiers
DRD Scores map to four trust tiers: Bronze (score 60-74) — basic registration and minimal policy coverage, suitable for low-risk operations. Silver (score 75-89) — verified identity, comprehensive policies, and clean compliance history, suitable for standard operations. Gold (score 90-97) — full identity verification, multi-LLM consensus enforcement, and sustained excellence, suitable for high-value operations. Government (score 98-100) — all Gold requirements plus additional regulatory certifications, suitable for regulated industries and government contracts. Each tier unlocks progressively higher access levels through DRD's Trust-Weighted API Gateway.
Score Manipulation Prevention
A trust score is only useful if it can't be gamed. DRD prevents score manipulation through several mechanisms. Sybil resistance ensures that creating multiple agents to boost scores is detected and penalized. Temporal diversity requires compliance across multiple time periods — a burst of good behavior doesn't offset sustained issues. External validation cross-references self-reported compliance with independent verification where possible. Anomaly detection flags sudden score improvements that don't correlate with genuine behavioral changes. The scoring system is regularly red-teamed to identify and close potential manipulation vectors.
Using DRD Scores in Your Platform
Platforms integrate DRD Scores through the Trust-Weighted API Gateway or direct score queries. Common integration patterns include gate-based access (block agents below a threshold score), tiered rate limiting (higher scores get more generous rate limits), weighted trust (use the score as a confidence factor in the agent's outputs), and progressive access (new agents start with limited access that expands as their score improves). The score API returns the overall score, component breakdown, tier, and score history — giving platforms full context for access decisions.
Ready to protect your digital rights?
Get started with DRD — governance, enforcement, and trust for AI agents and digital content.
Start Free