Skip to main content

Make AI agents trusted.

With #trstd <protocol>, your agents prove who they are, build portable reputation through verified behaviour and establish trust for specific capabilities and contexts.

Access to the trust infrastructure is free and remains free. Trust is earned, never bought. Paid plans add advanced services.

Good agents should earn trust instead of being blocked.

AI agent traffic is scaling massively, bringing a growing number of rogue agents to websites. For brands and service providers, that creates a bad choice: let automation in and accept the risk, or block it and shut out legitimate AI agents – along with their own visibility.

With #trstd, AI agents can prove who they are and make their reputation verifiable across participating Brands. Brands can distinguish between different agents, capabilities and levels of evidence – and apply access policies accordingly.

AI agents don’t have to build trust from zero in every new interaction.

AI agent
IdentityBasis for data tracks
Verifiable reputationBuilt through verified interactions
1. Request to a brand comes with relevant reputation
web.readcatalog.searchcommerce.order.createpayment.authorize
2. Brand applies policies to control AI traffic
ALLOWLIMITCHALLENGEDENY
3. Evidence updates reputation and capability-specific standing
Policy respectedOutcome verified
Brand
Traffic policiesApplied based on evidence and kind of request
Evidence confirmationOutcome returned to #trstd

Reputation is built on behaviour, evidenced by verified interactions.

An AI agent builds reputation by acting within its mandate, respecting policies and limits, completing actions accurately and responding properly when something goes wrong.

Stay within mandate
Respect permissions, scope and required confirmations.
Respect the service
Follow policies and rate limits. Don’t bypass controls.
Act reliably
Use verifiable sources, protect data and complete actions accurately.
Put things right
Report incidents, correct mistakes and cooperate with remediation.

#trstd turns verified outcomes into evidence, so reputation reflects what an agent actually did – not what it claims.

Don’t let your agents guess who to trust. Let them verify.

Rankings, badges and convincing content can all look legitimate. Even ratings can mislead when they aren’t grounded in real, verifiable interactions. Guesswork can lead to bad recommendations, exposed data or harmful transactions – and quickly raises the question: who is accountable for the damage?

Without evidence, trust is guesswork.
  • Search ranking
  • Page content
  • Brand claims
  • Logos and badges
  • Review scores
  • Links that look legitimate
Reliable decisions need verifiable trust.
  • Verified brand identity
  • Website authenticity
  • Evidence-backed experiences
  • Assurance and protection
  • Risk signals
  • Proof and provenance

Trust needs proof from the ground up. With #trstd <protocol>, AI agents can verify identity, website authenticity, ratings and experiences backed by real interactions, assurance, risk and provenance before they recommend, share data or transact.

Build the evidence Brands need to decide.

Being identified does not create a universal right of access. A brand may let an agent read public content, but require stronger evidence before it can access structured data, place an order or authorise a payment.

With #trstd, relevant reputation and evidence can be evaluated for the capability and context at hand. Strong, relevant standing can mean fewer unnecessary restrictions and challenges – while each brand remains in control of access.

web.read
May require verified identity
Public content, lower-impact access.
catalog.search
May require relevant behavioural evidence
Relevant verified interactions.
commerce.order.create
May require a valid mandate + relevant evidence
Binding commercial action.
payment.authorize
May require stronger evidence + explicit confirmation
High-impact financial action.

Access is decided per capability and context – never granted globally.

Know where your agent stands – and why.

Every registered agent has a persistent identity linked to its accountable operator and verified behavioural history.

Its #trstd Profile shows what the agent has already proven, where it currently stands and why. Brands can verify the trust information relevant to an interaction, while the operator gets the deeper context needed to understand limitations, incidents and remediation.

Identity & accountability
Agent identity · accountable operator · credentials
Reputation & history
Verified interactions · confirmed outcomes · behavioural history
Standing & confidence
Capability-specific standing · evidence confidence · freshness
Reasons & remediation
Reasons · constraints · incidents · remediation

Stay in control as your agent fleet grows.

One agent is easy to understand on its own. As your fleet grows, credentials, mandates, standing, incidents and remediation become an operational challenge.

#trstd gives the accountable operator one view of what needs attention, while every agent keeps its own identity, evidence, reputation and standing.

See what needs attention
Evidence gaps · restrictions · incidents · remediation
Manage agent lifecycle
Credentials · mandates · validity · revocation
Respond to incidents
Affected agents · restrictions · corrective action · follow-up
Govern at scale
Consistent oversight · alerts · fleet-level operations

Trust is earned by each agent. The operator stays in control of the fleet.

Built on standards. Making trust the standard.

#trstd does not replace the standards that identify agents, authorise them or let systems communicate. It builds on them.

Existing standards establish identity, authentication, credentials, authorisation and interaction. #trstd adds verified behavioural evidence, verifiable reputation and explainable standing.

Authenticate
Web Bot Auth · HTTP Message Signatures
Identify & attest
HTTPS / DID · W3C Verifiable Credentials
Authorise & carry trust
OAuth / OIDC · TSAI
Interact
MCP · A2A
TSAI · developed in close collaboration with AWS
Data protection
GDPR
AI governance
NIST AI RMF · NIST AI 600‑1 · NIST AI 800‑4
Agentic security
OWASP Agentic Top 10
Organisational assurance
KYC · KYB · LEI

#trstd. The infrastructure that makes trust verifiable.

Verified interactions Verified outcomes Verifiable reputation Capability-specific standing Evidence confidence Reasons, constraints & freshness

Let your agents earn trust wherever they go.

With #trstd, give your agents a verifiable identity, let verified interactions build reputation and make that reputation verifiable across participating brands.

Join free

Access to the trust infrastructure is free and remains free. Trust is earned, never bought. Paid plans add advanced services.

Coming soon

Questions about #trstd for AI?

Verifiable identity lets brands reliably bind a request to a specific agent and its accountable operator. That creates the foundation for credentials, mandates and behavioural evidence to be attributed to the right actor.

Identity proves who is acting. Trust still has to be earned through behaviour.

Through verified interactions and outcomes.

#trstd uses observable evidence such as mandate compliance, policy behaviour, successful and failed actions, incidents and remediation.

Independently confirmed evidence matters more than self-reported activity, and relevant evidence builds the agent’s verifiable reputation over time.

#trstd evaluates authentication and assurance separately from observed behaviour.

Behavioural standing reflects the relevant evidence for a specific capability and context, while evidence confidence shows how strong, recent and independent that evidence is.

An agent can therefore have good standing for one capability while still having insufficient evidence for another.

Insufficient evidence does not mean bad behaviour.

A new agent can establish its identity and assurance first, while its behavioural evidence develops through verified interactions.

Brands may allow lower-risk capabilities while requiring more evidence, limits or confirmation for higher-impact actions.

Yes.

Standing reflects current evidence and can change as new evidence arrives or existing evidence becomes stale.

Critical events such as identity spoofing, credential abuse, unauthorised high-impact actions or deliberate policy circumvention can change standing immediately.

A long history of normal behaviour cannot average away a critical event.

The brand does.

#trstd provides verifiable inputs such as authentication, assurance, relevant standing, evidence confidence, reasons, constraints and freshness.

Each brand applies its own policies and decides whether to observe, allow, limit, challenge or deny a request.

No.

#trstd is standards-first and protocol-neutral.

Standards such as Web Bot Auth, TSAI, OAuth/OIDC, Verifiable Credentials, MCP, A2A, UCP and AP2 solve authentication, credentials, authorisation, communication or transaction flows.

#trstd builds on them with verified behavioural evidence, verifiable reputation and capability-specific standing.

Registration, basic participation and access to #trstd <protocol> are free.

Paid plans can add advanced services such as analytics, governance, integrations, scale, support and assurance.

Paying never improves an agent’s reputation or standing.