AI Bias &
Governance Policy

Effective date: August 3, 2026

This document defines the systemic guardrails and internal audit metrics deployed by Waypoints Digital LLC to manage Artificial Intelligence integrations safely and responsibly within Waypoints Compass (“Compass”).

1. Approved Model Boundaries and Data Integrity

• Authorized Model Providers. Compass integrates two approved AI providers, each scoped to a specific function: (a) the Anthropic Claude API, the sole authorized Large Language Model for text generation in the Compass copywriting tools; and (b) the Google Gemini API, the sole authorized image model, used only for photo enhancement and artwork generation within the Create module. No other AI provider is integrated into Compass. Adding or substituting a provider requires a documented internal review and a corresponding update to this policy and to the Application Privacy Policy.
• Input Restrictions. Employees, system admins, and customers are strictly prohibited from inputting proprietary software source code, corporate financials, or guest personally identifiable information into text prompts. Photographs submitted to the image-enhancement feature are a deliberate and disclosed exception: a customer’s own property photographs may depict identifiable individuals, including guests. Customers represent that they hold the necessary rights and consents for any photograph they submit (see Master SaaS Subscription Agreement, Section 4); such photographs are processed transiently to produce the customer’s result; and Waypoints does not use them to construct any profile of a depicted individual.
• Data Training Opt-Out. Our API connections are explicitly configured to opt out of data sharing for model training purposes. Text prompts sent to the Anthropic API are handled strictly as transactional requests and are not utilized by Anthropic to train future iterations of their core models. Likewise, under our paid Google Gemini API terms, images and prompts transmitted for enhancement or artwork generation are not used by Google to train its models.

2. Mandatory Human-in-the-Loop Review

• Draft Classification. All social captions, ad copy, and marketing text variations generated via the Anthropic Claude API, and all enhanced photographs and generated artwork produced via the Google Gemini API, are classified strictly as unverified drafts.
• No Automated Publishing. The platform does not support or allow fully automated publishing pipelines from any AI model — text or image — to live ad networks or to published pages. All generated output requires manual review and approval by the campground owner or staff before deployment.
• Image Boundaries and Bias Controls. Artwork generation is limited to non-photorealistic graphics and must not be used to produce realistic depictions of real, identifiable individuals. Photo enhancement is constrained to image-quality attributes — lighting, straightening, clarity, and comparable corrections — and must not add, remove, or otherwise alter the substantive content of a scene. Operators are reminded at the point of generation that AI image models can reflect demographic and aesthetic bias, and that any imagery representing guests, staff, or the surrounding community should be reviewed for fair and accurate representation of the property as it actually is.

Interface Disclaimer Notices. The generation interfaces present the following disclosures to all operators.

Text generation: “All text variations created by this tool are AI-generated drafts. The user maintains sole responsibility to review, edit, and approve all copy for factual accuracy, bias, brand compliance, and legal safety before publishing.”

Image enhancement and artwork generation: “AI-enhanced photos and generated artwork are drafts. Enhancement may only improve image quality — it must never change what the scene actually shows. You are responsible for confirming that any image you publish honestly depicts your property, and that you hold the rights and consent for everyone pictured.”

3. Quality Controls and Review Protocols

Waypoints Digital LLC monitors the performance of the integrated Anthropic Claude and Google Gemini APIs using a risk-based approach. Rather than automated loops, our engineering team reviews system prompt effectiveness and output trends periodically based on operational needs. To facilitate this, Compass features an inline user flagging utility covering both text and image output. When a subscriber flags an output as inaccurate, biased, or otherwise inappropriate, the originating prompt — and, for images, the enhancement parameters applied — is isolated for review, allowing our team to tune our system instructions and generation constraints.

← All policies