AI Policy

How CiteLadder uses AI: answer engines generate responses, scoring is deterministic, and Customer Data is never sold or used to train third-party models.

Last updated ·

Overview

Cube27 IT Private Limited operates CiteLadder, an AEO analysis product. We observe how answer engines describe brands and products, persist raw responses as evidence, and score them with deterministic rules.

This AI Policy summarises how AI systems are involved. It does not replace our Privacy Policy at /privacy or our Terms of Service at /terms.

How AI is used

  • Answer engines (ChatGPT, Gemini and Claude) generate responses when you run audits. Those calls use your BYOK credentials where configured. Google AI Overviews are read from Google search results through DataForSEO, not generated by a model.
  • Scoring of mentions, citations, and related visibility metrics is deterministic over persisted artifacts — not an LLM judging another model’s answer.
  • Optional product features may use models for assistance (for example drafting or research helpers). The in-app Agent always uses platform providers (Anthropic or OpenAI), not customer-supplied keys. When a feature uses a model, we will describe the purpose in-product.
  • Generated prompt suggestions and internal link suggestions are checked by a quality-judging model that answers fixed yes-or-no and multiple-choice questions. It ranks and filters suggestions for your review; it does not write them, publish them, or score your business.
  • A visibility result records what a provider returned for one execution. It does not establish that the answer is true, permanent, or what another user would see.

What we do not do

  • We do not sell Customer Data.
  • We do not use Customer Data to train third-party foundation models.
  • We do not fabricate scores when evidence is missing — unavailable metrics render as an em dash.
  • Generation alone never publishes content, makes a purchase from CiteLadder, or changes an external system.

Human oversight

CiteLadder is a measurement and evidence tool for professional teams. Review generated recommendations and drafts before relying on or publishing them. You remain responsible for decisions you make using the outputs. Raw answers and rule versions are available so you can verify scores.

CiteLadder does not guarantee a search ranking, citation, AI recommendation, traffic increase or commercial result. Provider models, data and availability can change independently of CiteLadder.

AI outputs and third-party data are provided as received, for your own evaluation, and are not professional advice. You are responsible for how you, your users and any AI assistant you connect through the API or MCP use them. Cube27 is not liable for decisions, publications or other use based on them, or for their misuse. Our Terms of Service at /terms set out these limits in full.

Contact

Questions about this Policy: contact@cube27.com.