About
AI Content & Data Policy
How we use AI assistance, what stays human-verified, and how we disclose AI authorship to readers and search engines. Plus our aggregator stance, Egypt PDPL and GDPR compliance, and data-source attribution.
What we use AI for
Gouna Realty operates as an aggregator across multiple Egypt real-estate listing platforms. We index thousands of properties in El Gouna, structure the data, and present the catalogue with our own editorial layer. AI assistance is part of that layer.
We use AI tools to draft property descriptions when the source platform provides only a thin or formulaic listing. The AI generates an expanded description grounded in the structured fields (bedrooms, bathrooms, square metres, neighborhood, features, view type, year built, price). It does not add facts that are not present in the source data.
We also use AI assistance for area guides, neighborhood overviews, and buyer-process explainers. These pieces are drafted with AI support, then reviewed for factual accuracy, local relevance, and tone.
What stays purely human-written
Pricing data, exact location coordinates, agency contact details, square-metre figures, and amenity lists are sourced directly from the original listing platform. No AI rewrites these fields. When a price changes on the source, our scraper updates the value within hours. The AI-drafted description is not regenerated unless the underlying structured data changes materially.
Founder commentary, market analyses with named authors, and any advice on the buyer process is written by our editorial team and bylined accordingly.
Which AI model we use
We use Anthropic Claude (Haiku and Sonnet model families) for text drafting. The choice reflects Claude's track record on grounded, source-anchored writing and our internal review of output quality across multiple AI providers.
Model versions and provider mix can change as the AI tooling landscape evolves. The disclosure principle does not change: wherever AI assistance is used, we say so.
How we review AI-drafted content
Every AI-drafted property description passes a factual-claim check before publication. The reviewer compares the description against the structured property data and flags any sentence that cannot be supported by the underlying fields. Flagged sentences are rewritten or removed. The reviewer also strips marketing superlatives ("stunning", "breathtaking", "hidden gem") that are not grounded in concrete evidence.
Area guides and buyer-process explainers undergo a second pass for local accuracy. Claims about El Gouna infrastructure, transportation, schools, hospital access, golf-course details, and visa procedure are checked against authoritative sources (Orascom Development Holding documentation, Egypt FRA filings, Egypt Ministry of Interior visa pages, and on-the-ground reports from our El Gouna network).
How we mark AI-assisted content
The site-wide footer carries a one-line disclosure on every page. Individual property pages with AI-drafted descriptions additionally show a small badge near the description heading indicating that the text is AI-enhanced and human-verified.
Structured data (Schema.org JSON-LD) embedded in our pages includes a creator field that names the organization responsible for the AI assistance and an editor field naming the human reviewer where applicable. This makes our AI authorship pattern legible to search engines and AI assistants.
The llms.txt and llms-full.txt files we publish for AI agents include an explicit notice that property descriptions and area guides may be AI-assisted, with the human-review process described.
Why we disclose
Google's spam policies require disclosure when AI is used to produce content at scale. Transparency about AI authorship is acceptable to Google and to readers, provided the content is factually accurate and human-reviewed. Hiding AI assistance is both a policy violation and a trust violation. We disclose by default.
Disclosure is also the right answer for our buyers. Anyone making a property decision should know which information is structured source data, which is editorial commentary, and which is AI-drafted text that has been factually verified. Our disclosure pattern makes those layers visible.
Aggregator, not brokerage
Gouna Realty is an aggregator-platform. We do not broker deals. Listings are republished from six verified source platforms with full attribution and back-links to the originals. The six sources are Property Finder, Bayut, Aqarmap, Nawy, Dubizzle, and Gouna360. Each listing keeps its source-platform name and source-URL field so every claim is traceable.
Phase 1 (current): one platform-agency entity, Gouna Realty Platform. Buyer-direct agency partnerships open in Phase 2 during 2026. Until then, qualified inquiries route via WhatsApp to the listing-agency that owns the source-record. The agency handles the viewing, the negotiation, and the contract.
Egypt PDPL compliance
Gouna Realty complies with Egypt's Personal Data Protection Law (Law 151/2020, PDPL). Data-subject rights are active: access, correction, and deletion can be requested by emailing info@gouna-realty.com. Requests are answered within 30 days per PDPL Article 10.
Cross-border transfers to processors outside Egypt follow PDPL Chapter 4 with standard contractual safeguards. No personal data is sold or shared with third parties outside the lead-routing flow described above.
EU GDPR compliance
For visitors in the European Union, processing relies on EU GDPR Article 6.1.f (legitimate interest in operating the aggregator) and Article 6.1.a (consent) for cookies that are not strictly necessary. Cookie consent is granular and withdrawable at any time via the cookie banner.
Lawful-basis statements per processing activity, retention periods, and our Data Protection Officer contact are published in the full Privacy Policy.
AI-content discipline
AI-content generation follows three hard rules: schema-validation runs pre-publish so every JSON-LD block parses against Schema.org; no fabricated property data — numeric fields mirror the structured source record; every numeric claim on persona-routes is sourced to the public llms.txt baseline published at gouna-realty.com/llms.txt.
Questions or corrections
If you spot a factual error in any AI-drafted description or area guide, message us on WhatsApp or email info@gouna-realty.com. We correct flagged content within 48 hours and revise the review checklist if a recurring issue surfaces.
Last revised: June 7, 2026. About Gouna Realty