Skip to content
Transparency

How we use artificial intelligence

Cementum uses artificial intelligence to read, cross-check and cite sources. Here we explain what it does, what it does not do and how to recognise generated content.

This is a courtesy translation: in case of any discrepancy, the Italian text prevails.

1. What artificial intelligence does in Cementum

Cementum uses artificial intelligence models from third-party providers to read plans, implementing regulations, building regulations, constraints and maps, to cross-check them and to write a report in which every statement links to its source. The report proposes a verdict (feasible, feasible with conditions, not feasible) and states for each result how solid it is: Verified, To confirm, Gap, Not expressed.

2. What it does not do

  • It does not decide for you: the decision always stays with the person.
  • It does not replace a qualified professional and it is not a legal opinion, an appraisal, a signed drawing or a building application.
  • It does not certify or approve: it maps the situation and cites the sources, so you can check them.
  • It does not hide what it has not found: whatever could not be confirmed is stated.

3. Human oversight

Every result comes with its source, which you can open and read. In the Pocket Urbanist service the case is handled by a selected expert from the Cementum team, who checks the results at the source before delivery and explains them on a call. When a report has been reviewed by an expert, it carries the «Reviewed by a professional» badge.

4. When you interact with an artificial intelligence system

In the Cementum app, the answers to your questions are generated by an artificial intelligence system, and the app says so. This site does not use chatbots: messages sent through the contact and waiting-list forms are read by people on the team.

5. How to recognise generated content

  • App reports. Reports and exported PDFs state that they are generated with artificial intelligence.
  • Published cases. Case documents are produced by an expert from the team, who leads the case from start to delivery and uses artificial intelligence as support; we say so at the top of every case.
  • Images. The photorealistic renders in the documents are images generated with artificial intelligence: they are neither photographs nor approved designs. On the site, every page that contains them carries the label «Contains an AI-generated image».
  • Marking in the files. The page images, documents and preview images of the cases also contain a machine-readable marking (IPTC and XMP metadata on the digital source type) indicating the use of artificial intelligence.
  • 3D model. The 3D study model is a reconstruction from the project drawings: it is neither a survey nor a photograph.

6. Limits and errors

Like any research tool, Cementum can make mistakes: a document may be out of date, a source may not respond, a regulation may be read incompletely. That is why every statement is cited and we ask you to have the results checked by a qualified professional before acting. If you find an error, write to us at info@cementum.ai.

7. Your data

The data and documents you upload are used to work on your project. Which providers process them, where and for how long is described in the privacy notice.

8. Team training

Everyone at Cementum who develops the platform or handles cases with artificial intelligence is trained on how it works, its limits and how to verify sources.

9. References and contact

This page meets the transparency obligations of Regulation (EU) 2024/1689 on artificial intelligence (Article 50, applicable from 2 August 2026) and the promotion of AI literacy (Article 4). For questions: legal@cementum.ai.