AI Policy

Citely is an AI-native product: we query third-party AI engines to measure your visibility, and we use AI models internally to generate content drafts and audit recommendations. This page explains how.

Tracking third-party AI engines

To measure visibility, we send tracked prompts to AI engines and record their public responses. Every plan covers ChatGPT, Gemini and Perplexity; Claude is included from the Pro plan and Grok on the Scale plan. This uses each provider’s standard API access and does not involve your private data — only the prompts and public responses relevant to your tracked topics.

Recommendation generation

Weekly fix recommendations, generated files (sitemap, llms.txt, agents.md, schema), and plain-English narratives are generated using large-language-model providers under commercial agreements that exclude your content from being used to train the providers’ general-purpose models.

Human review is expected

AI-generated recommendations, generated files, and audit findings are decision-support, not final output. Review AI-generated content for accuracy before publishing — like any AI system, outputs can contain errors or reflect stale information.

Your data and model training

We do not use your account data, uploaded content, or tracked visibility results to train models for other customers. Aggregate, de-identified benchmarking data (for example, category-level visibility trends) may be used to improve product recommendations.