Responsible AI

01Our commitment

Globaleur builds AI infrastructure for travel enterprises. Because our systems influence recommendations and service experiences for real travelers, we hold ourselves to a standard of responsible, accountable, and transparent AI. This statement describes the principles and practices that guide how we design, deploy, and govern our technology.

02Principles

Our work is guided by five principles:

• Human-centered. AI augments people and operators; it does not replace human judgment in consequential decisions.
• Fair. We work to identify and reduce unjust bias in our systems.
• Transparent. We are clear about where and how AI is used.
• Private & secure. We protect traveler data by design.
• Accountable. We govern, monitor, and take responsibility for our systems.

03Human oversight

Our platform is designed to keep humans in control. Customers configure guardrails, escalation paths, and approval steps for sensitive actions, and operators can review, override, or disable automated behavior. We do not design systems intended to make legally or materially significant decisions about individuals without human review.

04Fairness & non-discrimination

We assess models and recommendation logic for unjust disparities and provide controls that help customers apply consistent, policy-aligned treatment across traveler segments. We do not knowingly build systems that discriminate on protected characteristics, and we provide documentation to support customers’ own fairness assessments.

05Transparency & explainability

We help customers disclose AI involvement to travelers where appropriate and provide observability into how recommendations are produced — including the signals considered and policy constraints applied — so decisions can be reviewed and audited.

06Privacy & data minimization

We process only the data needed to deliver the Services, apply tenant isolation, and support retention and deletion controls. Customer conversations and traveler data are not used to train shared foundation models without explicit authorization. See our Privacy Policy and Data Processing Addendum.

07Safety & abuse prevention

We implement guardrails to reduce harmful, deceptive, or unsafe outputs, monitor for misuse, and provide mechanisms for customers to constrain model behavior to their policies. We support model-agnostic routing so customers can select models that meet their safety and compliance requirements.

08Model governance & evaluation

We maintain governance processes covering model selection, evaluation, change management, and incident response. Material changes to model behavior are tested before release, and we track quality and safety metrics over time.

09Customer responsibilities

Responsible outcomes are a shared effort. Customers are responsible for configuring the platform consistent with their legal obligations, providing appropriate notices to travelers, validating outputs for their context, and maintaining human oversight of consequential actions.

10Reporting concerns

If you believe a Globaleur-powered system has produced a harmful, unfair, or unsafe outcome, contact us at legal@globaleur.com. We investigate reports and use them to improve our systems and safeguards.