The 167 AI Signals — Complete Reference for B2B
What Are the 167 AI Signals?
The 167 AI signals are the atomic machine-readable data points that AI systems — including ChatGPT, Gemini, Claude, Copilot, and Perplexity — use to discover, classify, verify, and recommend businesses. They are not content. They are structured infrastructure. They exist in JSON files, schema markup, linked data graphs, and edge-delivered KV records — not in the website's visible text.
Every signal has a specific function. Every signal belongs to one of three categories. Together, the 167 signals form the complete machine-readable representation of a business in the AI knowledge graph.
A business with all 167 signals correctly deployed is, from an AI perspective, a completely legible entity — findable, understandable, verifiable, trustable, and recommendable. A business with no signals deployed is invisible — regardless of how good its website looks or how long it has been operating.
The Three Signal Categories
Identity and Entity
Signals that establish who the business is — name, category, location, legal identifiers, and entity relationships in the knowledge graph.
Deep Dive →Trust and Proof
Signals that prove the business is who it claims to be — certifications, citations, cryptographic proof, and external verification references.
Deep Dive →Intent and Governance
Signals that map the business to buyer intent and declare its governance posture — service categories, buyer personas, and compliance declarations.
Deep Dive →The Four Signal Status Tags
AI AUDIT™ assesses every signal and assigns one of four status tags. The status map is delivered as signal-assessment.json — consumed directly by AI SIGNALS™ for systematic deployment.
| Status | Meaning | Action Required |
|---|---|---|
| 🟢 Present | Signal exists, correctly structured, accessible to AI crawlers | None — monitor with AI SONAR™ |
| 🔴 Missing | Signal does not exist in any accessible location | Deploy with AI SIGNALS™ |
| 🟡 Inconsistent | Signal exists but conflicts with data in another source | Resolve conflict, redeploy |
| ⚪ Unverifiable | Signal exists but cannot be confirmed from external source | Add external reference or proof |
Category Deep Dives
Identity and Entity Signals
48 signals · Prerequisite categoryThese 48 signals establish the business's identity in the AI knowledge graph. They cover legal name, all name variants, trading name, primary category, subcategory, NACE/SIC code, founding date, founding location, current headquarters, registered address, all operating locations, service geography, canonical URL, primary contact, secondary contacts, all social profile URLs, company registration number, VAT number, entity type, entity status, parent entity, subsidiary entities, and all sameAs references to authoritative external profiles.
Identity and Entity signals are the prerequisite category. AI cannot process Trust or Intent signals for a business it cannot identify. These 48 signals must be deployed first.
Read the complete 48-signal reference →Trust and Proof Signals
61 signals · Highest weight for recommendationThese 61 signals prove the business is who it claims to be and that its claims are accurate. They cover all certification references with issuing body URLs, all award references, all professional body memberships, case study count and schema status, client testimonial schema, third-party review schema with aggregate rating, citation count, citation quality classification, sector directory presence, authoritative directory presence, sameAs link validity, external mention consistency, SHA-256 proof file status, OpenTimestamps anchor status, proof file accessibility, evidence registry status, third-party verification status, verification badge deployment, all governance file statuses, HITL protocol documentation, EU AI Act readiness declaration, GDPR compliance declaration, and all transparency declarations.
Trust and Proof signals carry the highest weight in AI recommendation decisions. A business with all 61 deployed is treated as an authoritative source. Without them, even a perfectly Findable™ business will not be recommended.
Read the complete 61-signal reference →Intent and Governance Signals
58 signals · Buyer matching layerThese 58 signals map the business to buyer intent and declare its governance posture. They cover primary service category, all secondary service categories, all service-to-intent mappings, primary buyer persona, all secondary buyer personas, buyer journey stage coverage, primary problem solved, all secondary problems solved, primary solution approach, all delivery format declarations, pricing model, pricing range, geographic intent filters, sector intent filters, company size intent filters, role intent filters, language intent filters, all negative intent declarations, AI crawler permission declarations, llms.txt status, ai.json status, intents.json status, ai-actions.json status, ai-routes.json status, allow-lane-matrix.json status, and all governance and compliance file statuses.
Without Intent signals, AI systems cannot match the business to the correct buyer queries. A business with no intent mapping will appear for generic queries — if at all — and miss the specific buyer intent queries where conversion probability is highest.
Read the complete 58-signal reference →The Signal Deployment Pipeline
Signals follow the canonical ADI™ rule: Signals → Files. Never Files → Signals. Signal values are determined from the Business Knowledge Graph™ first. Files are generated from those values. Reversing the order produces inconsistent data.
AI LENS™ — Build the Knowledge Graph
Extract and structure all business knowledge into the canonical knowledge base
AI SIGNALS™ — Generate Signals
Generate all 167 signals from the Business Knowledge Graph™
AI SIGNALS™ — Validate
Four-stage validation: completeness, consistency, verifiability, accessibility
SIGNAL COMPILER™ — Compile Files
Generate all signal files: ai.json, entities.json, intents.json, and seven others
EDGE INJECTOR™ — Deploy to Edge
Serve all signal files from Cloudflare Workers KV — sub-10ms globally
AI SONAR™ — Monitor
Continuously monitor signal health and AI visibility across all five systems
Signal Coverage = AI Visibility Score
The AI Readiness Score produced by AI AUDIT™ is directly derived from signal coverage. Identity and Entity signals contribute 35% of the score. Trust and Proof signals contribute 45%. Intent and Governance signals contribute 20%. A business with all 167 signals deployed and validated scores in the 90–100 range — the Recommendable™ threshold.
How Signals Are Weighted
Not all signals carry equal weight in AI recommendation decisions. The Trust and Proof category — 61 signals — has the highest combined weight because it directly determines whether AI systems treat a business as a credible, authoritative source. The specific high-weight signals within this category are:
- External citations — Very High. References from independent sources carry the most weight.
- Third-party verification — Very High. eu-ai-audit.eu badge is the strongest single trust signal.
- SHA-256 + OTS proof — High. Cryptographic proof makes claims independently verifiable.
- Professional body membership — High. Strongest sector authority signal for regulated industries.
- Case study schema with outcomes — High. Proves real-world authority with measurable results.
- sameAs link validity — Medium-High. Enables cross-source entity verification.
Frequently Asked Questions About the 167 AI Signals
What are the 167 AI signals?
The 167 AI signals are the atomic machine-readable data points that AI systems use to discover, classify, verify, and recommend businesses. They fall into three categories: Identity and Entity (48), Trust and Proof (61), and Intent and Governance (58).
Why are there exactly 167 signals?
The 167 signals represent the complete set of structured data points that AI systems currently use to evaluate business entities. The number reflects the union of signal requirements across all five major AI systems — ChatGPT, Gemini, Claude, Copilot, and Perplexity.
Which signal category is most important?
All three categories are required. However, Identity and Entity signals are the prerequisite — AI cannot process Trust or Intent signals for a business it cannot identify. Trust and Proof signals have the highest weight for recommendation frequency.
How are signals deployed?
Signals are deployed via the ADI™ pipeline: generated from the Business Knowledge Graph™ by AI SIGNALS™, compiled into structured files, and served from the Cloudflare edge by EDGE INJECTOR™. The canonical rule: Signals → Files. Never Files → Signals.
What happens if signals are inconsistent across sources?
Inconsistent signals fragment the business's knowledge graph entry and reduce AI confidence. AI systems weight consistent signals higher than inconsistent ones. AI AUDIT™ identifies all inconsistencies in its 167-signal status report.