Trusted
The Trusted pillar is about evidence. Every claim in an Echo can be traced back to proof, so AI agents can explain their recommendations.
The challenge
When an AI agent recommends a venue, a guest might ask "why this one?"
The agent needs more than "because I said so." Evidence provides the answer: sources describing an intimate atmosphere, verification from a regional tourism board, consistent information across platforms.
What Trusted contains
Evidence
The proof behind claims. Sources, provenance, confidence scores.
Verification
Third-party validation. DMO credentials, verifiable proofs, trusted issuers.
Convergence
Multiple sources agreeing. Cross-validation that builds confidence.
An example
A boutique hotel claims an "intimate atmosphere." The Trusted pillar backs this up:
| Layer | What it shows |
|---|---|
| Evidence | Multiple sources describe intimate, romantic, peaceful |
| Provenance | Data drawn from multiple platforms |
| Confidence | High score based on volume and agreement |
| Verification | VisitScotland Quality Assured credential |
When an agent recommends this hotel for a romantic anniversary, it can explain why with specifics.
What agents can do
With trust data, agents can:
- Explain recommendations with specifics rather than assertions
- Compare venues based on evidence strength
- Flag uncertainty when sources disagree
- Prefer verified venues when confidence matters
The structure makes AI reasoning transparent rather than opaque.