# Defensive measures against misinformation

German title: Defensive Maßnahmen gegen Falschinformationen

> False statements about a brand in AI answers can be addressed in five steps: document, check sources, publish your own details unambiguously, have third-party sources corrected, report errors to the provider. That these steps change the answer is not verified.

## Facts

- **Approach:** working model in five steps
- **Google reporting:** thumbs down, “Report a problem”
- **Other providers' reporting:** not verified
- **Effect:** not verified
- **Legal questions:** case by case, legal advice

## The five steps (recommendation)

- Document: record answer, prompt, system, settings (e.g. logged in, with web search) and date.
- Check sources: if the answer shows sources, compare their content with the statement.
- Make own details unambiguous: a public page with verified details, tables and lists instead of platitudes. Structured data must match the visible content and does not guarantee rich results. Check whether the providers' search bots may fetch the page (see Brand retrieval and RAG ).
- Correct third-party sources: ask operators of comparison sites, forums and review portals for corrections, with evidence.
- Report and recheck: report errors to the provider and repeat the same question after some time.

## Reporting channels

- Google AI Overviews: Below each overview there is a thumbs up and thumbs down. With thumbs down choose “Report a problem”, pick a category and add details if possible. The feedback includes the last search query and the results. Alternatively use the three-dot menu of the overview.
- OpenAI, Perplexity, Anthropic: The documentation read contains no information on reporting channels for wrong answers. They were not verified. An effect on later answers is not verified.

## Limits

That false statements mostly trace back to a few faulty sources is not verified; a language model can also produce false statements without a faulty source. Whether a statement is unlawful is decided case by case. This report makes no statement on this; for business-damaging falsehoods, seek legal advice.

## Note on origin

This report combines the earlier reports on misinformation and on competitor hijacking. The evidence base is the same for both: the steps are a working model.

## Not verified

This report describes a working model of the author. There is no primary source from Google, OpenAI, Perplexity or Anthropic for the individual steps; read them as recommendations, not as verified effects. Only the reporting channel for Google AI Overviews is verified.

## Sources

- Google Search Central: Intro to structured data (primary source)
- Google Search Help: Feedback zu AI Overviews (primary source)
- OpenAI: Overview of OpenAI crawlers (primary source)
- Perplexity Docs: Crawlers (primary source)
- Anthropic Support: Does Anthropic crawl data from the web? (primary source)

Checked on: 24 October 2026.

## Questions

### Why do classic SEO methods help little here?

That is not verified. Whether the problem lies with third-party sources or the model must be checked case by case.

### What is a fact sheet?

A public page with verified details on a product or organization. The term is a convention, not a standard.

### How do I report a wrong AI Overview to Google?

Below the overview click thumbs down, then “Report a problem”, choose a category, add details and “Submit”. An effect on later overviews is not verified.

### Is a thumbs down in ChatGPT enough?

The documentation read makes no statement on this. An effect on later answers is not verified.

### Does a cease-and-desist letter help?

That depends on the case and must be clarified legally. This report gives no legal advice.

## Related

- [Third-party platforms and sentiment: what is verified](https://munichmedia.com/reports/aeo-sentiment-drittplattformen.md)
- [Brand retrieval & RAG: what can be influenced](https://munichmedia.com/reports/aeo-brand-retrieval-rag.md)
- [Schema.org for AEO: what is verified](https://munichmedia.com/reports/aeo-schema-org-basis.md)

Canonical HTML: https://munichmedia.com/reports/aeo-defensive-massnahmen