The four stages of evidence
Every trust signal on the platform sits at exactly one of these stages, and the stage is always visible.
- Stage 1
Declared
Stated by the participant and not yet supported by anything inspectable.
- Produced by
- The participant, in their own words.
- What you can rely on
- Treat as context only. Declared statements never affect ranking or trust badges.
- Stage 2
Evidenced
Backed by an inspectable artefact — a published AI deployment, case study, milestone or uploaded delivery evidence.
- Produced by
- The participant, through the deployment, case study or delivery lifecycle.
- What you can rely on
- You can read the artefact yourself and judge it on its detail.
- Stage 3
Verified
Reviewed against the standard by AI Deploy Network before it was allowed to be published.
- Produced by
- Platform review, through the verification and publication workflow.
- What you can rely on
- The artefact passed review. Verification covers authenticity, not commercial fit.
- Stage 4
Confirmed outcome
A measured business result — hours automated, cost saved or revenue impacted — recorded against completed implementation delivery.
- Produced by
- Implementation delivery, with evidence attached to the outcome.
- What you can rely on
- The strongest signal on the platform. Confirmed outcomes carry the most weight.
How strong is a single piece of evidence?
Every AI Deployment and AI Case Study carries exactly one of three labels. They are progressively stronger: each one adds a check the one before it does not make.
- Level 1
Builder Reported
Supplied by the builder and not independently verified by AI Deploy Network or a client.
Produced by: The builder who submitted the record.
- Level 2
AI Deploy Network Approved
The record passed AI Deploy Network's review before publication. Approval covers the record itself, not an independently confirmed business outcome.
Produced by: AI Deploy Network moderation.
- Level 3
Engagement Verified
The work traces back to a delivery outcome the client organisation confirmed on this platform.
Produced by: The client organisation, through engagement confirmation.
AI Deploy Network Approved means the record passed our review — it does not mean the business outcome has been independently confirmed. Only Engagement Verified evidence is traced to a delivery outcome the client organisation confirmed on this platform, with their publication consent still active. Builder Reported information has not been independently verified. Builder verification badges are a separate, identity-level signal and never imply that a record is Engagement Verified.
One badge vocabulary
10 badges, five families. A badge means the same thing on a profile, a deployment, a case study, a recommendation and a benchmark.
Identity and standing
- Communitydeclared
The account exists and is in good standing. No evidence has been reviewed yet.
Appears on: Builder profiles, Marketplace cards
- Verifiedverified
Identity and professional claims were reviewed by AI Deploy Network before publication.
Appears on: Builder profiles, Marketplace cards, Recommendations
- Endorsedverified
Verified, and additionally endorsed by AI Deploy Labs after direct review of delivered work.
Appears on: Builder profiles
- Builder organisationverified
An organisation verified to run implementation delivery end to end, not only to advise on it.
Appears on: Builder organisation profiles
Delivered evidence
- Published deploymentverified
An AI deployment that passed review and is published as inspectable evidence of delivered work.
Appears on: AI deployments, Builder profiles
- Published case studyverified
A reviewed implementation story with a stated situation, approach, result and supporting detail.
Appears on: AI case studies, Builder profiles
- Delivery evidenceevidenced
Documents, reports, metrics or links attached to a delivery milestone by the people doing the work.
Appears on: Opportunity delivery, Milestones
Business outcomes
- Confirmed outcomeconfirmed
A measured business result recorded against completed delivery — hours automated, cost saved or revenue impacted.
Appears on: Builder profiles, Case studies, Recommendations, Benchmarks
Explainability
- Explained resultevidenced
Every ranked or recommended result carries the reasons it appeared, in the order they were applied. No black-box scores.
Appears on: Marketplace results, Discovery, Recommendations
Privacy
- Privacy gatedverified
Aggregated comparisons are published only when enough independent contributors exist and no single contributor dominates the result.
Appears on: Benchmarks
The principles behind the standard
Evidence, not claims
A statement never becomes a trust signal on its own. It becomes one when there is an artefact behind it that you can open and read.
Explainable, not black box
Rankings, recommendations and readiness decisions state their reasons. If a result cannot be explained, it is not shown as a trust signal.
Confirmed outcomes carry the most weight
Measured business results from completed delivery outrank profile completeness, activity or self-description everywhere they are used.
Privacy is engineered, not promised
Aggregate comparisons pass a privacy gate before publication: a minimum contributor count, a dominance ceiling and outlier suppression.
One vocabulary, everywhere
The same badge means the same thing on a profile, a deployment, a case study, a recommendation and a benchmark.
What the standard does not assert
A standard is only useful if its boundaries are stated as plainly as its claims.
- Verification confirms that work is authentic and reviewed. It is not a guarantee of commercial performance.
- Confirmed outcomes are reported against a specific implementation and context; they are not a forecast for yours.
- Benchmarks describe aggregated peers, never an individual organisation or participant.
- The platform does not sell placement. Ranking position cannot be purchased.
See the standard in practice
Every ranked result, published deployment and confirmed outcome on the platform is produced by the stages above.