Loading...
Loading...
Zero-Knowledge Proofs
Prove compliance to regulators, auditors, and federation members without revealing a single byte of proprietary data. DRD uses zero-knowledge proofs to make trust verifiable and privacy absolute.
How It Works
Zero-knowledge proofs let you convince any verifier that a statement is true without revealing any information beyond the validity of the statement itself.
Private Witness
Your data stays local
Proof Generation
snarkjs Groth16 / PLONK
Public Verification
Anyone can verify
Select a pre-built compliance circuit for your regulatory framework, or define custom constraints. Circuits encode the exact criteria that must be satisfied.
The DRD SDK generates a zero-knowledge proof on your infrastructure. Your private data serves as the witness — it never leaves your environment.
Share the compact proof with regulators, auditors, or federation members. They verify it in milliseconds using only the public verification key — no access to your data needed.
Capabilities
Generate compliance proofs locally using your private data. Only the cryptographic proof is shared — the underlying records, models, and configurations remain completely private.
Generate verifiable proofs for EU AI Act, NIST AI RMF, ISO 42001, and custom frameworks. Each proof cryptographically attests to specific compliance criteria without revealing how you meet them.
Prove your DRD Trust Score exceeds a required threshold without revealing the exact score. Partners and regulators verify the proof on-chain — no trust required.
Federation members verify each other's compliance status using ZKPs. No member ever sees another's proprietary data, training sets, or internal governance policies.
Built on Groth16 and PLONK proving systems via snarkjs. Audited circuits with trusted setup ceremonies ensure proof soundness and zero-knowledge guarantees.
All proofs are non-interactive — generate once, verify anywhere, any number of times. No back-and-forth protocol required between prover and verifier.
Proof Types
DRD ships pre-built circuits for the most common compliance verification scenarios. Custom circuits are supported for enterprise and federation use cases.
Prove adherence to a regulatory framework without exposing internal policies, model weights, or training data.
Prove your trust score, risk rating, or compliance percentage exceeds a minimum without revealing the exact value.
Satisfy data sovereignty requirements by proving compliance to foreign regulators without transferring any data across jurisdictions.
Prove your audit logs are complete and unmodified without revealing the log contents or the actions they record.
Developer Integration
import { DRD } from '@drd/sdk';
const drd = new DRD({ token: 'drd_live_sk_...' });
// Generate a zero-knowledge compliance proof
const proof = await drd.zkp.generateComplianceProof({
agentId: 'agent_123',
framework: 'eu-ai-act',
});
// Proof contains only the cryptographic attestation
// — no private data, no model details, no training info
console.log(proof.proofId); // "zkp_abc123..."
console.log(proof.circuit); // "eu-ai-act-v2"
console.log(proof.publicSignals); // [1] (compliant = true)
// Anyone can verify with just the proof ID
const valid = await drd.zkp.verify(proof.proofId);
console.log(valid); // true
// Generate a trust score threshold proof
const scoreProof = await drd.zkp.generateThresholdProof({
agentId: 'agent_123',
metric: 'drd-trust-score',
threshold: 85,
// Proves score >= 85 without revealing exact value
});
// Federation members verify each other's proofs
const federationValid = await drd.zkp.verifyFederationProof({
proofId: scoreProof.proofId,
federationId: 'fed_eu_governance',
});Zero-knowledge proofs let you satisfy every regulator, every auditor, and every federation partner — without ever exposing your proprietary data.