AI Usage Disclosure

We believe in being upfront about how AI is used inside GeoGibbon, what it does well, and where it has limits. This page explains both — so you know what you are getting and can make informed decisions about acting on the audit.

1. We use AI to audit AI

GeoGibbon's whole purpose is to test how generative AI platforms describe your business. To do that we send queries to:

• OpenAI (ChatGPT and search-enabled variants)
• Anthropic (Claude with web search)
• Perplexity
• Google (AI Overviews and Gemini, where accessible)

These platforms produce the answers we analyse. We do not control what they generate, and their output can change from week to week as they update their models, training data, and search backends.

2. We use AI to draft analysis

Our audits combine deterministic data (citation counts, scoring matrices, technical scans of your site) with AI-assisted writing of recommendations and prose. For the Free Mini and Standard tiers, recommendations are drafted with AI assistance and reviewed by automated quality checks. For the Expert tier, every recommendation is reviewed by a human GeoGibbon strategist before delivery.

3. What this means: AI can be wrong

Generative AI sometimes hallucinates — invents facts, misattributes quotes, or generates plausible-sounding but incorrect claims. We use prompting techniques and validation layers to minimize this, but we cannot guarantee zero errors.

You should sanity-check our recommendations against your own knowledge of your business and market before implementing them. If anything in your audit looks wrong or surprising, contact us — we will investigate and correct.

4. What we do NOT do with AI

• We do not let AI decide alone whether to charge you for re-runs or refunds — those are human decisions.
• We do not generate fake reviews, testimonials, or case studies for your business.
• We do not use AI to write reviews or comments on third-party platforms (Google reviews, TripAdvisor, etc.) on your behalf.
• We do not recommend deceptive content tactics.

5. What data we send to AI providers

For each audit, we send the AI providers:

• Queries about your business and your market (e.g., "best Italian restaurant in Bangkok")
• Excerpts of your public website content for analysis
• Excerpts of public information about your competitors

We do not send AI providers:

• Your email address
• Payment information
• Personal details about you or your team beyond what is publicly listed on your business website

6. AI providers' data retention

The AI providers we work with each have their own retention policies. When possible, we use API endpoints with zero-retention or short-retention defaults. However, we cannot guarantee how each provider handles data on their end.

For the most up-to-date information, see:

• OpenAI's privacy policy
• Anthropic's privacy policy
• Perplexity's privacy policy

7. Audit results are a snapshot

AI search behavior is in flux. The audit you receive reflects what the platforms answer at the moment we run the queries. A query that does not mention you today may mention you next month, and vice versa. This is why all paid tiers include re-scans — they give you a moving picture instead of a single frame.

8. Have questions?

If you want to know more about how we use AI in your specific audit, or have concerns about data handling, email us at [email protected] and we will respond directly.