Technology & AI Policy

We Are Guests uses AI-assisted automation to keep up with fast-moving travel news. This page explains exactly how — what the system does, what it draws on, what safeguards apply, and the firm line between automation-assisted news and our own first-hand reviews.

Last updated 17 May 2026

Transparency by Default

We Are Guests is a small, family-run publisher covering a large and fast-moving subject — theme parks, cruises, resorts, and travel deals across several destinations. To keep pace with travel news without cutting corners on accuracy, we use an AI-assisted newsroom automation system. We would rather explain that openly than have readers wonder.

This page covers how the automation works and what safeguards apply. The short version, and how to read the byline on any given article, is on the Authenticity page.

What the Automation Does

Our newsroom automation handles travel news — and only news. It runs as a pipeline that discovers, drafts, checks, and publishes coverage of developments across the destinations we follow.

Core capabilities:

  • Automated discovery from a curated set of travel sources every day
  • AI-assisted drafting using established large language models
  • A multi-pass quality-control stage that checks claims against source material
  • Human editorial review before anything is published

Why we built it

Comprehensive coverage of theme-park changes, cruise-line announcements, route and pricing updates, and travel deals would take a newsroom far larger than two people. Rather than cover a fraction of it by hand, we built automation to handle the volume so that the human team can focus on what only people can do: visiting places, forming opinions, and exercising editorial judgement.

Where Automation Is Never Used

Our reviews, comparisons, and destination guides are written entirely by us, from our own visits and our own opinions. Automation plays no part in them. Specifically, AI does not:

  • Write or score hotel, resort, restaurant, cruise, or attraction reviews
  • Decide our opinions or recommendations
  • Invent first-hand experience
  • Produce opinion, analysis, or commentary without a named human author
  • Decide what is worth covering, or how sensitive a story should be handled

Automation-assisted news carries the We Are Guests Newsroom byline. Reviews and guides carry our own names. The byline always tells you which is which.

Source Material

The automation monitors a curated set of credible travel sources — operator announcements from parks, resorts, and cruise lines; official tourism information; verified deal and offer feeds from booking partners; and established travel-news outlets with their own editorial standards.

Every claim in a published news article must be traceable to a specific piece of source material the system drew from. The automation does not add detail that did not come from a named source.

Models and Providers

The system uses commercial large-language-model services from established providers, chosen per pipeline stage for capability, speed, and cost. Lightweight steps run on faster, cheaper models; drafting and verification run on stronger ones.

We do not pin specific model versions on this page, because models change faster than policy documents. The standing commitment is that any model we use must meet our internal bar for accuracy and faithfulness to source material. If we change provider or model in a way that materially affects the published product, we will say so here.

Accuracy and Fairness Safeguards

The drafting stage is instructed against editorialising, sensationalism, speculation, and hype. Headlines are held to the same standard as body text. Beyond that:

  • Generated coverage is checked against its source material before publication
  • Prices, dates, and offer terms are verified against the operator or booking partner
  • Reader-reported concerns about coverage are reviewed by a human editor
  • Where a pattern of error appears, the pipeline is adjusted — not patched article by article

Your Data

We collect minimal reader data:

  • Standard, aggregated web analytics
  • Email addresses only where a reader opts in

What we do not do:

  • We do not use reader data to train AI models
  • We do not sell reader information to third parties
  • We do not share reader data with AI providers beyond what is operationally necessary

The automation works from published travel sources, not from reader data. The full privacy posture is on the privacy policy page.

Limitations

Verification at the quality-control stage can confirm that a news article's claims are supported by the source material it drew from. It cannot catch an error in the underlying source itself, and it does not replace the editorial judgement that decides whether a story is worth running. When automation falls short, a human editor steps in. When an error reaches publication, we correct it under our corrections policy.

Questions

Questions or concerns about how we use automation can be sent to editorial@weareguests.net. We respond to every substantive message.

This policy is part of our wider standards framework. See also: Authenticity, Trust, Editorial Standards, and Corrections.