The Fides Methodology.
A decentralized framework for turning unstructured information into verified, immutable assets. Our core methodology (Source-Reshape-Value) ensures data integrity in the age of AI.
Data flows through a three-stage validation pipeline before being minted as a trusted asset.
1. Ingestion
Raw data is fetched from global media nodes. Metadata is extracted and origin is cryptographically signed.
2. Provenance Validation
The Fides Trust Function checks the source's transparency markers and signs the data's origin.
3. Minting
Verified content is structured into a JSON-LD Asset, hashed, and stored in the Registry as an immutable record.
The Fides Trust Score (T) is not subjective. It is calculated via a weighted algorithm that penalizes opacity and rewards consistency.
The percentage of past claims verified as true by consensus nodes over a rolling 12-month window.
A score (0-1) based on ownership disclosure, citation density, and clear correction policies.
Verification depth. Tier AAA sources require 1-node confirmation; Tier B requires 3+ nodes.
Algorithmic deduction for detected emotive language or undisclosed conflict of interest.
To ensure decentralization, we distinguish betweenauditable logic (Open Source) and secure execution (Protected).
●Open Source (Public)
- Trust Function Formulas (T(s))
- Weighting Coefficients (w1, w2, w3)
- Consensus Threshold Rules
- API Documentation & SDK
●Protected (Secure)
- Node Private Signing Keys
- Anti-Sybil Defense Patterns
- Internal DDoS Mitigation Logic
- Validator Anonymity Layers
Consensus Node Voting & ZKP Flow
*Zero-Knowledge Proofs (ZKP) allow nodes to verify source metadata without exposing sensitive publisher credentials.
The Fides network is maintained by a diverse group of stakeholders who provide computational power and reputational backing.
Foundational Nodes
Academic & Research Inst.Participate for public good and data integrity. They gain early access to verified global datasets for academic research.
Strategic Nodes
Media & Tech PartnersEntities with "Skin in the game." They verify to protect their own IP and prevent AI-driven misinformation.
Governance Nodes
Standards CouncilValidators who ensure cross-border compliance. Incentivized by governance rights and protocol influence.
Once verified, information is standardized into theFides Asset Object (FAO). This structure allows AI models to consume trusted data programmatically.
{
"@context": "https://fidesstandard.com/ns/v1",
"asset_id": "FIDES-2025-X892-A",
"timestamp": "2025-05-12T08:30:00Z",
"trust_score": 98.5,
"grade": "AAA",
"source": {
"name": "Reuters",
"fides_id": "FS-SRC-004"
},
"content_hash": "0x7f83b1657ff1fc53b92dc18148a1d65dfc2d4b1fa3d677284addd200126d9069",
"verification_proof": {
"method": "node_consensus",
"validators": ["FS-NODE-01", "FS-NODE-09"]
}
}The content_hash is a one-way mathematical derivative. It is impossible to reconstruct the original article or identify individual authors from the hash alone. This ensures the protocol remains GDPR Compliant while providing absolute proof of existence.
The methodology applies differently depending on your role in the ecosystem.
Verification Standard
For Verifiers & Consumers
Rules for assessing external information. Detecting fabrication, deepfakes, and calculating the T score.
Read Standard A →Production Standard
For Publishers & Creators
Rules for creating trusted content. How to structure metadata to be "Fides-Ready" at the point of origin.
Read Standard B →