Artificial intelligence is reshaping property valuations in South Africa by combining trusted data, machine learning, and human expertise to produce faster and more accurate results, moving beyond valuation as professional judgment and historical comparables.
The scale and depth of data underlying the use of AiVM, Lightstone’s AI-powered tool used for residential property valuation, matters because valuation models are only as strong as the information they learn from.
“By combining deeds records, cadastral data, municipal rolls, listing data, point-of-interest information, and imagery, Lightstone’s system builds a much richer view of property value than traditional methods alone”, Hayley Ivins-Downes, Managing Executive Real Estate, Lightstone Property, said at The Africa Valuation Conference 2026 recently.
In a presentation titled “Our Journey and Application of AI in AIVM”, Ivins-Downes outlined how Lightstone’s data includes nationwide coverage of registered South African properties, millions of indexed properties and street addresses, and decades of deeds transaction history.
The data asset
Ivins-Downes said AIVM used features such as property location, size, previous sale price, and other structured inputs to estimate market value.
To make the output useful for decision-makers, AIVM includes a confidence score and a safety score, helping users understand not only the estimate itself but also how much trust to place in it. That emphasis on explainability was important because it kept the model transparent rather than turning it into a black box.
The outcome? It offers market insights which track property value inflation, monitors property transfer activity and days on market, highlights percentage of asking prices attained across regions, and merges proprietary data across property, automotive, and retail sectors to create competitive advantages.
Ivins-Downes emphasised that AI should support, not replace, human experts, and she highlighted “human-in-the-loop” decision-making, where the model assisted analysts, valuers, and lenders while still leaving room for professional oversight.
Ivins-Downes said Lightstone was empowering the property and financial ecosystem through explainable intelligence, around three pillars:
- Transparency: Unlike generic “black box” systems, Lightstone’s AI “white box advantage” provided predictions that can be interrogated and understood by human decision-makers.
- Integrity: Lightstone is Africa’s only member of the European AVM Alliance and met strict global benchmarks for statistical rigor and independence. In addition, Lightstone’s “Narrow AI” is trained on South African nuances, interpreting deeds logs, municipal data gaps, and informal market anomalies that generic models miss.
- Outcomes: Include banking efficiency (instant “lending-grade” market values at the offer-to-purchase stage, leading to higher deal approval rates and reduced friction in bond origination), and strategic precision.
Lightstone is also exploring computer vision, satellite imagery, and other advanced techniques to improve coverage and accuracy.
Overall, Ivins-Downes said AI is a practical tool for the property and financial sectors, not just a technical experiment. Its message is that the future of valuation depends on trustworthy data, continuous model improvement, and the ability to explain results clearly.
