….Safe AI.Travel


Part of an IBM Partner Programme. Association Partnership Programme

Ethically Sourced™ Safe AI For Your Members

Verifiable AI Agent Identities • Provenance • Accountability • Continuous Assurance

Introducing the Ethically Sourced™ AI Agent Programme for Travel, Tourism and Hospitality

SafeAI.TravelAgenticAI.TravelEthically Sourced™ MarkTAER AssuranceTravel.1Hospitality.1

The Challenge of AI Agents in Travel & Tourism

AI Agents are becoming active participants in the travel industry — discovering suppliers, recommending services, communicating with businesses, and increasingly initiating commercial interactions.

The Fundamental Question

How can a traveller, travel business, or another AI Agent establish who is behind an AI Agent, where it comes from, what it is authorised to do, and whether its assurance status remains valid?

SafeAI.Travel is developing a framework to address this challenge through verifiable AI Agent identities, the Ethically Sourced™ mark, and ongoing assurance and authority controls supported by TAER (Team Agent Emergency Revocation).

Introducing the Ethically Sourced™ Mark

A recognisable mark backed by verifiable evidence. The Ethically Sourced™ mark provides a visible indication that an AI Agent has satisfied defined requirements under five essential assurance dimensions:

Dimension Assurance Criterion & Required Evidence
1. Identity A persistent, machine-readable identity through which the AI Agent can be identified and its associated records resolved.
2. Provenance Evidence of the Agent’s origin, relevant software or model lineage, deployment arrangements, and applicable provenance information.
3. Authority Evidence identifying the actions the Agent is permitted to perform, who granted those permissions, and applicable operational restrictions.
4. Operational Controls Evidence of relevant safeguards, including access controls, continuous monitoring, incident handling, change management, and revocation arrangements.
5. Accountable Operator A legal entity or responsible organisation operating the AI Agent and accountable for its activities.

Ethically Sourced™ indicates defined requirements have been assessed against verifiable evidence.

TAER: Assurance That Can Be Withdrawn

Trust Should Remain Verifiable

Under the proposed framework, an AI Agent’s Ethically Sourced™ status is linked to a machine-readable assurance record detailing:

  • The AI Agent and its accountable operator
  • Applicable assessment criteria and evidence trails
  • Scope, limitations, assessment dates, and review requirements
  • Current status: Active • Suspended • Expired • Withdrawn
  • References to relevant authority and revocation information

Where criteria cease to be met, TAER supports suspension or withdrawal of the mark authorisation. Effective operational withdrawal requires enforcement systems to recognise and act upon revocation.

The 6-Stage Recognition Lifecycle

01

REGISTER

Establish the Agent’s persistent identity and link it to its accountable operator.

02

RESOLVE

Publish appropriate machine-readable identity records and Canonical AI Endpoint references.

03

VERIFY

Examine evidence relating to identity, provenance, operator accountability, and claimed authority.

04

ASSESS

Apply agreed TAER methodology to relevant operational safeguards and programme requirements.

05

RECOGNISE

Authorise use of the Ethically Sourced™ mark and publish a verifiable status record.

06

MONITOR & REVOKE

Maintain review, suspension, and withdrawal processes so status stays current as circumstances change.

Association Member Services & Architecture

A New Member Service for Tourism and Travel Associations

Tourism and travel trade associations are ideally positioned to help members prepare for AI Agent discovery and commerce:

  • AI Agent Identity Provision: Persistent, resolvable identities linked to accountable operators, credentials, capabilities, and Canonical AI Endpoints.
  • Ethically Sourced™ Mark Assessment: Structured route for evaluating eligible members’ AI Agents.
  • TAER Assessment & Status Management: Periodic review, evidence evaluations, and suspension/withdrawal mechanisms.
  • Web2 and Web3 Discovery: Conventional internet discovery alongside optional Handshake pathways (including Travel.1 and Hospitality.1).
  • Agent Authority Management: Support for recording, checking, and withdrawing delegated permissions within participating systems.
  • Education & Member Readiness: Practical guidance on identity, provenance, safety, accountability, and responsible deployment.

The Role of AgenticAI.Travel

AgenticAI.Travel provides the coordinated machine-readable identity and discovery architecture behind the assurance programme:

Persistent Identifiers → Canonical AI Endpoints → Operator Records → Authority Credentials → Web2 / Web3 Pathways (Travel.1 / Hospitality.1)

Summary Principle: The name enables discovery. The identity establishes the reference. The evidence supports verification. TAER manages assessment and status. The mark communicates assurance.

Proposed 90-Day Founding Pilot

We are inviting selected tourism and travel trade associations to help shape a practical member programme:

Pilot Phase Operational Focus & Deliverables
Phase 1: Association Consultation Agree member needs, relevant AI Agent use cases, assurance criteria, and programme responsibilities.
Phase 2: Member Onboarding Establish pilot Agent identities.
Phase 3: Evidence & TAER Assessment Collect and assess applicable identity, provenance, authority, operational-control, and accountable-operator evidence.
Phase 4: Ethically Sourced™ Demonstration Demonstrate how an eligible Agent’s mark authorisation can be issued, independently checked, suspended, and withdrawn.
Phase 5: Association Programme Review Evaluate results, member feedback, governance arrangements, and potential for a wider member service rollout.

Why This Matters to Your Members

Establish AccountabilityDemonstrate ProvenanceClarify AuthorityCommunicate AssuranceRetain ControlPrepare for AI Commerce

3–5 November 2026 | ExCeL London

Meet Us at WTM London 2026

As part of an IBM Partner Programme, we are introducing a Travel AI identity initiative and proposed SafeAI.Travel Association programme around World Travel Market London 2026.

We welcome discussions with national tourism associations, hotel associations, travel agent and tour operator associations, destination organisations, travel technology associations, and professional membership bodies. Appointments arranged with an IBM Partner to move things forward.

SafeAI.Travel • AgenticAI.Travel • Ethically Sourced™