By Fritz Swanepoel, CEO of Leapfrog Property Group 

Ask the average person how the real estate industry is using artificial intelligence (AI), and they’ll most likely talk about agents using large language models (LLMs) to artificially stage properties. However, there’s so much more to AI than LLMs, and, when applied properly, it is able to transform the industry in ways that make it fairer for everyone.

It could aid in dismantling generational inequalities in how we buy, selland occupy property. Used ethically, AI can help democratise our property market.

There is no denying that our housing market remains entangled with historical legacy and human subjectivity. For decades, buyers of colour and individuals from marginalised backgrounds have navigated a property landscape rife with invisible friction. Even when a buyer has the financial means to purchase a home in an affluent neighbourhood and secure a bond, the human element can unintentionally introduce barriers. This is rarely intentional. Instead, it comes from hidden blind spots in the industry that manifest in quiet ways, such as overly strict screening or property valuations based on who lives in the area rather than market data.

This is where AI can help by replacing subjectivity with objective facts. Take property valuations as an example. These have historically been vulnerable to local preconceptions. When AI assists in valuing a home, it focuses on data such as recent local sales, municipal zoning, deeds office records and the building’s actual condition. As a result, you are likely to get a valuation based on market reality that doesn’t see the identity of the buyer or seller. This allows estate agents to provide clients with neutral, data-backed guidance.

This fairness also applies to how buyers find homes. Even the most well-meaning agents can occasionally make subconscious assumptions about what a client wants or where they would feel comfortable living, accidentally filtering out great properties. AI-powered matching engines remove this human error. By focusing strictly on what the buyer asks for, like price, suburb preferences, and size, the algorithm makes sure everyone sees the same listings. A buyer’s background is removed from the equation, making the search open and available to everyone based on financial fit.

Beyond matching buyers to homes, AI also has the potential to uncover opportunities that are often overlooked by the market. By analysing thousands of data points simultaneously, including infrastructure investment, school performance, transport development, municipal spending, crime trends and transaction activity, AI can identify suburbs that may be poised for growth long before those trends become obvious. This gives buyers, investors and estate agents access to deeper market intelligence, helping them make decisions based on emerging realities rather than outdated perceptions or anecdotal opinions.

Screening buyers and tenants is another area ripe for change. For years, vetting has frequently relied on subjective impressions, which may introduce bias. AI screening shifts the focus to verified financial indicators. Even better, AI can look at non-traditional data to make assessments more inclusive. For the millions of self-employed individuals and entrepreneurs in South Africa who lack a traditional corporate credit history, AI can analyse consistent rental payments, utility bills and cash flow patterns. This creates a fairer picture of financial reliability in a fraction of the time.

This could be particularly transformative in South Africa, where millions of economically active people earn income outside of traditional employment structures. Small business owners, entrepreneurs, freelancers and members of the informal economy often struggle to demonstrate their financial reliability through conventional credit assessments. AI has the potential to broaden access to property ownership by evaluating a wider range of financial behaviours and payment patterns, allowing responsible buyers who may previously have been overlooked to participate more fully in the property market.

However, adopting AI brings ethical risks that we cannot afford to ignore. We have come to learn that technology is never inherently neutral, especially not AI. It gives us back the data we feed it because AI models draw on decades-old lending and housing records. This means that it can copy and amplify the systemic discrimination we want to eliminate. If old records are full of past inequalities, an unchecked AI will label certain neighbourhoods or groups as high-risk, disguising old biases as hard math.

Compounding this is the “black box” problem. Some advanced AI models are so intricate that even their developers cannot fully explain the logic behind a specific outcome. If an AI system declines a bond application or undervalues a property, the lack of transparency makes it difficult for a client to challenge the decision or understand how to address it. For estate agents, this makes it hard to stand by the technology and goes against our promise of fairness and consumer protection.

To overcome these challenges, we cannot simply deploy AI and walk away. As estate agents, lenders and industry stakeholders, we must actively govern these systems by regularly testing for bias, ensuring transparency and being able to explain decisions to consumers in plain language. Ultimately, fairness does not depend on how sophisticated the technology is, but on how responsibly it is managed.

The greatest contribution AI may make to the South African property market is not faster transactions, automated marketing or greater efficiency. It may be its ability to remove the hidden assumptions that have influenced property decisions for generations. If we combine the analytical power of AI with strong human oversight and ethical leadership, we have an opportunity to build a property market that is not only smarter, but more accessible, transparent and inclusive for all South Africans.