2. Ingredients System

At the heart of AIOX is the Ingredients System — a method of breaking down digital content into atomic semantic units that are machine-readable, verifiable, and richly described.

2.1 What is an Ingredient?

An ingredient is the smallest meaningful unit of content made visible to AI. Unlike brittle Q/A pairs, ingredients are rich, structured, and semantically tagged.

2.2 JSON Schema Example

Below is an illustrative JSON schema for a text ingredient, showing metadata, provenance, and licensing.

{
  "ingredient_id": "ing-7f3a9b12",
  "version": "1.0",
  "type": "text",
  "content": {
    "text": "To change a flat tire, first secure the vehicle on level ground..."
  },
  "metadata": {
    "intent": "instructional",
    "tone": "neutral",
    "utility": "how-to",
    "target_audience": "general",
    "psychographic_tags": ["drivers", "car_owners"]
  },
  "provenance": {
    "author": "John Doe",
    "publisher": "ExampleAutoBlog",
    "hash": "c82e5c45d9f1a5e07c7c3a8d67abfe8d...",
    "signature": "ed25519:9a1c7b1e...f43c2"
  },
  "licensing": {
    "training": "allowed",
    "citation_required": true,
    "expiration": "2026-09-22"
  }
}

2.3 Multi-Modal Support

Ingredients are not limited to text. AIOX supports Image, Audio, Video, and Dataset ingredients.

{
  "type": "image",
  "content": { "url": "...", "alt_text": "A car jack lifting a sedan" },
  "metadata": { "intent": "illustrative", "narrative_type": "photo" },
  "licensing": { "training": "denied" }
}

{
  "type": "dataset",
  "content": {
    "schema": { "columns": ["Country", "Average Tire Lifespan (km)"] },
    "data": [ ["USA", 60000], ["India", 40000] ]
  }
}

3. Capsules

If ingredients are the atoms, Capsules are the molecules. A capsule is a structured package that groups related ingredients into a coherent whole (e.g., an article, a product page, a podcast episode).

3.1 Structure of a Capsule

A capsule acts like a digital "box" that AI systems can open. It includes the content (ingredients), rules of use, version history, and proofs of authenticity.

3.2 Example: News Article Capsule

{
  "capsule_id": "caps-news-ev-2025",
  "version": "1.0",
  "ingredients": [
    "ing-text-001", // Headline
    "ing-text-002", // Body paragraph
    "ing-img-101",  // Chart image
    "ing-vid-201"   // Analysis video
  ],
  "context": {
    "narrative_order": ["ing-text-001", "ing-text-002", "ing-img-101"],
    "topic": "Electric Vehicles Market",
    "summary": "European EV sales surpassed gasoline cars in Q3 2025."
  },
  "provenance": {
    "publisher": "GlobalNews",
    "hash": "root987...",
    "signature": "ed25519:xyz..."
  },
  "licensing": {
    "training": "allowed",
    "citation_required": true
  }
}

4. Manifests

A Manifest (typically aiox.jsonc) is the machine-readable index that serves as the entry point for AI crawlers. It functions like a sitemap.xml but for the AI era.

4.1 Example Manifest

{
  "aiox_version": "1.0",
  "publisher": {
    "name": "GlobalEnergyInstitute",
    "website": "https://globalenergy.org",
    "public_key": "ed25519:abcdef123456..."
  },
  "default_licensing": {
    "training": "allowed",
    "citation_required": true
  },
  "capsules": [
    {
      "id": "caps-dataset-2025",
      "url": "https://.../caps-2025.json",
      "hash": "rootdataset123..."
    },
    {
      "id": "caps-article-policy",
      "url": "https://.../caps-policy.json",
      "hash": "rootarticle456..."
    }
  ],
  "delta_feed": {
    "url": "https://globalenergy.org/aiox/delta-feed.json",
    "last_updated": "2025-09-22T10:30:00Z"
  }
}

5. Delta Feeds

Delta Feeds are incremental update streams that tell AI systems exactly what has changed since their last crawl. Instead of re-ingesting a full dataset, a crawler only fetches the differences.

Why It Matters

6. Cryptographic Verification

AIOX builds a Chain of Trust using standard cryptographic primitives. This ensures that content ingested by AI is authentic, untampered, and legally attributable.

7. Licensing & Governance

AIOX embeds licensing controls directly into every layer (ingredient, capsule, manifest), ensuring that AI systems cannot ignore them. It transforms ingestion from a lawless scrape into a rights-aware protocol.

7.1 The Framework

• Training Permissions: Allow, deny, or restrict AI model training.
• Citation Rules: Require attribution in AI responses.
• Expiration: Content can "age out" of training datasets.
• Monetization: Gate access via subscription or pay-per-use APIs.

7.2 Example: Subscription Access

{
  "capsule_id": "caps-ft-article-0925",
  "licensing": {
    "training": "restricted",
    "citation_required": true,
    "monetization": {
      "subscription_required": true,
      "api_endpoint": "https://ft.com/aiox-api/validate"
    }
  }
}

In this example, a crawler must validate its subscription via the API endpoint before ingesting the full content.

8. Fallback Exposure

Not all crawlers will be AIOX-compliant immediately. To ensure visibility during the transition, AIOX provides redundant fallback layers.

9. Analytics & Visibility

Today's publishers are blind to AI consumption. AIOX makes every AI interaction measurable, structured, and actionable.

10. Ecosystem & Adoption

AIOX is not just a specification — it is the foundation of a global ecosystem.

11. Integrations

AIOX offers turnkey integrations for major Content Management Systems, allowing you to publish AI-optimized capsules without writing a single line of code.

Platform Roadmap