AI Reputation

Correction request

The AI Reputation report reports what the generative engines publicly say — it does not endorse them. If an answer about you is inaccurate, here's how to act, on both sides.

On the Levier IA side

We can annotate your report to flag contested information, and remove a clearly erroneous verbatim from your diagnostic. Write to us with the elements below — no form needed:

  • your brand name and your official website;
  • the engine(s) concerned;
  • the information to correct, and the accurate information with an official source;
  • the link to your AI Reputation report, if you have it.

On the engine publishers' side

Levier IA does not host the AI answers and cannot change what these engines generate. A durable fix goes through the publishers themselves as well as your own public sources:

  • Reinforce your official sources — rich pages (about, offers, legal notices), schema.org structured data, and an llms.txt file allowing AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended). This is the most effective lever: the engines rely on reliable third-party sources that cite you by name.
  • Report the error to the publisher via the feedback channels of ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Perplexity and Grok (xAI).

Your rights

If an answer contains inaccurate personal data about you, you have in particular a right to rectification and to object (GDPR, art. 16 and 21), exercisable with the relevant data controller — the engine's publisher. This page is informational and does not constitute legal advice.

See the methodology · Run a diagnostic