4 Protection Layers

Content Protection

DRD provides four layers of content protection — from passive fingerprinting to cryptographic provenance verification. Each layer adds depth to your defense.

Fingerprinting

Watermarking

Registry

Verification

Layer 1

Passive Fingerprinting

Content fingerprinting creates a unique hash of your content that can detect unauthorized copies even when the content has been modified.

VideoSHA-256 per-frame hashing

Cryptographic frame hashing for tamper detection and content identification

ImagesdHash perceptual hashing

Near-duplicate detection even after crops and filters

TextMinHash + locality-sensitive hashing

Document similarity detection across paraphrases

Layer 2

Active Watermarking

Invisible watermarks embedded in your content that survive compression, cropping, and format conversion.

Video SealNeural watermarking

Deep learning-based watermark embedding for video

VideoShieldRobust detection

Watermark extraction even after heavy processing

VideoMarkLightweight embedding

Low-overhead watermark for high-volume pipelines

Layer 3

Provenance Registry

Cryptographic proof of content origin using industry standards.

C2PA v2.2Content Credentials

Signed manifest proving who created content, when, and with what tools

Merkle TreeTamper-evident anchoring

Proof that content existed at a specific point in time

Layer 4

AI Pipeline Verification

For AI-generated content, DRD verifies the complete pipeline from training data to inference output.

Training DataProvenance verification

Verify data sources and licenses used in model training

Model CheckpointsCryptographic identity

Tamper-proof model versioning and integrity checks

Inference AttestationOutput verification

Proof that specific output came from specific model + input

Content Scanning Pipeline

When content is submitted for scanning, it passes through all four protection layers in sequence. The pipeline returns a unified match result with confidence scores from each layer.

content-scan.json

POST /api/v1/content/scan

{
  "contentUrl": "https://example.com/suspect-image.png",
  "contentType": "image",
  "contentHash": "sha256:f1e2d3c4b5a6...",
  "layers": ["fingerprint", "watermark", "provenance", "ai-pipeline"],
  "threshold": 0.85
}

// Response
{
  "ok": true,
  "data": {
    "scanId": "019scan-abcd-...",
    "matchCount": 2,
    "layers": {
      "fingerprint": {
        "matched": true,
        "similarity": 0.94,
        "algorithm": "dHash"
      },
      "watermark": {
        "matched": true,
        "watermarkId": "wm-019...",
        "confidence": 0.98
      },
      "provenance": {
        "matched": true,
        "c2paValid": true,
        "originalRegistration": "019content-..."
      },
      "aiPipeline": {
        "matched": false,
        "trainingSetDetected": false
      }
    },
    "matches": [
      {
        "contentId": "019content-abcd-...",
        "title": "Original Photograph - Sunset Beach",
        "similarity": 0.94,
        "owner": "Acme Photography Inc."
      }
    ],
    "scanTimeMs": 342
  }
}

Perceptual Hashing Algorithms

Unlike cryptographic hashes that change with any byte modification, perceptual hashes capture the visual or auditory essence of content. Two images that look the same to humans produce similar pHash values even after editing.

Algorithm	Content Type	Robustness
dHash	Images	Resize, crop, compress, color shift
frame-temporal-v1	Video	Re-encode, frame drop, speed change
chromaprint-v3	Audio	Transcode, clip, volume normalize
simhash-lsh	Text	Paraphrase (partial), truncation, reorder

C2PA Content Credentials

DRD signs content with C2PA Content Credentials (v2.2), creating a tamper-evident provenance chain. Credentials include the creator identity, creation timestamp, editing history, and DRD verification status.

c2pa-verify.json

POST /api/v1/content/verify

{
  "contentUrl": "https://cdn.example.com/verified-image.jpg"
}

// Response
{
  "ok": true,
  "data": {
    "valid": true,
    "manifest": {
      "title": "Original Photograph - Sunset Beach",
      "format": "image/jpeg",
      "claimGenerator": "DRD.io/1.0",
      "signatureInfo": {
        "issuer": "DRD Platform CA",
        "algorithm": "Ed25519",
        "time": "2026-02-10T15:30:00Z"
      }
    },
    "provenance": [
      { "action": "c2pa.created", "when": "2026-02-10T15:30:00Z", "softwareAgent": "DRD.io" },
      { "action": "c2pa.edited", "when": "2026-02-11T09:00:00Z", "softwareAgent": "Adobe Photoshop" }
    ],
    "tamperEvidence": "none_detected"
  }
}

DMCA Automation

The DMCA automation pipeline handles the full lifecycle of takedown requests, from initial detection through enforcement and resolution tracking.

Detect

Content scanning pipeline identifies potential infringement. Match confidence and evidence are collected automatically.

Notify

DMCA notice is generated from a legally-reviewed template. Sent to the hosting platform via their abuse API or email.

Track

Each takedown request is tracked in the enforcement dashboard with status updates and response timelines.

Enforce

If the platform does not respond within the statutory period, escalation actions are triggered (legal counsel, counter-notice).

Evidence Collection & Legal Export

DRD automatically collects and packages evidence for legal proceedings. Evidence packs include timestamped screenshots, fingerprint match data, watermark extraction results, and chain of custody documentation.

evidence-pack.json

GET /api/v1/content/:contentId/evidence-pack

{
  "ok": true,
  "data": {
    "contentId": "019content-abcd-...",
    "generatedAt": "2026-02-13T12:00:00Z",
    "format": "pdf+json",
    "evidence": {
      "registration": {
        "timestamp": "2026-01-10T08:00:00Z",
        "chainHash": "sha256:a1b2c3..."
      },
      "fingerprints": {
        "phash": "d4e5f6...",
        "matchSimilarity": 0.94
      },
      "watermark": {
        "extracted": true,
        "watermarkId": "wm-019..."
      },
      "chainOfCustody": [
        { "action": "registered", "at": "2026-01-10T08:00:00Z" },
        { "action": "match_detected", "at": "2026-02-12T14:00:00Z" },
        { "action": "evidence_collected", "at": "2026-02-12T14:01:00Z" }
      ]
    },
    "downloadUrl": "https://api.drd.io/evidence/019content-abcd.zip"
  }
}