Send a case for careful review

The lab studies how generative AI systems retrieve, cite, confuse and omit Italian companies, brands and places. Its focus is identity: name, place, branch, category and source support. It looks at generated answers as evidence trails rather than as simple rankings.

The lab does not run advertising campaigns, manage search accounts or promise placement in AI answers. It also does not treat one prompt result as a complete diagnosis. Its work is observational, comparative and focused on public evidence.

Read each note as a worked example of how an answer may have been assembled. The useful part is often the method: what was recorded, which source supported which claim, and where the identity began to drift.

Yes. The lab welcomes topic suggestions when they involve Italian businesses, brands, places, categories or bilingual answer paths. A strong suggestion includes the prompt used, the generated answer, the business identity in question and any public sources that seem relevant.

The lab replies when a request fits its research focus and can be assessed from public evidence. Some messages may receive only a short scoping response. The team avoids private claims that cannot be checked against public sources.

The lab answers in batches rather than as an instant support desk. Straightforward messages usually receive a shorter reply first, while complex case suggestions may need a slower review before the team can say whether the topic fits.

The site is available in English and Italian. The lab compares Italian and English answer paths because both can shape how Italian businesses are retrieved, cited and described in generated answers.