EU AI Act for Real Estate & PropTech: Transparency, Valuations & When AI Turns High-Risk

PropTech and real-estate firms use AI for automated valuations, property search, lead qualification and virtual staging. Under the EU AI Act (Regulation (EU) 2024/1689), the bulk of these uses are limited- or minimal-risk, with Art. 50 transparency duties for chatbots and generated media. The clearest escalation point is mortgage lending: AI that assesses a borrower's creditworthiness is high-risk under Annex III.

Is it in scope?

General proptech AI — automated valuation models (AVMs), search ranking, virtual staging and listing generation — is not listed in Annex III and is typically minimal-risk, while customer chatbots and AI-generated imagery attract Art. 50 transparency duties. The material trigger is Annex III(5)(b): AI used to evaluate the creditworthiness or credit score of consumers — directly relevant to mortgage and financing decisions — is high-risk. Where AI is used to determine access to social or public housing benefits, Annex III(5)(a) may also apply.

Typical AI use cases

  • Automated valuation models (AVMs) for property pricing
  • Property search and recommendation ranking
  • Lead qualification and sales chatbots
  • Virtual staging and AI-generated listing imagery
  • Mortgage creditworthiness assessment

Risk classification

For most agencies and portals the classification is limited- or minimal-risk: disclose AI chatbots and mark AI-generated imagery under Art. 50 (from 2 August 2026), with AVMs and search ranking generally minimal-risk. The escalation is specific — AI performing mortgage or financing creditworthiness assessment is high-risk under Annex III(5)(b), triggering the full provider or deployer obligations for that feature. AI determining access to public housing support may engage Annex III(5)(a). Property valuation of assets, as opposed to scoring people, does not itself trigger the credit listing.

Obligations to prepare for

AI-interaction disclosure for chatbots (Art. 50)
Marking of AI-generated imagery (Art. 50)
AI literacy of staff (Art. 4)
Assess mortgage credit scoring for Annex III(5)(b)
Human oversight where a feature is high-risk (Art. 14)
GPAI downstream provider/deployer duties

Your exact duties also depend on whether you build or use the AI. See obligations by operator role — provider, deployer, importer, distributor or GPAI provider.

FAQ

Is our automated valuation model high-risk?

Generally no. AVMs price property assets rather than score people, so they are not caught by the Annex III(5)(b) creditworthiness listing and are typically minimal-risk. Consumer credit scoring for a mortgage, by contrast, is high-risk.

When does proptech AI become high-risk?

The clearest case is mortgage and financing decisions: AI that evaluates the creditworthiness or credit score of a natural person is high-risk under Annex III(5)(b). AI determining access to public housing benefits may also engage Annex III(5)(a).

Do we have duties for our property chatbot and AI-staged photos?

Yes. Under Art. 50 you must disclose that a chatbot is an AI system, unless obvious from context, and mark AI-generated or manipulated imagery as artificial. These transparency duties apply from 2 August 2026.

See exactly what applies to your system

Run the free 2-minute risk checker — no sign-up.

Run the free check