Home buying made easy: Can AI really help?

Introduction: AI’s potential in the home buying process

On average it takes around 15 weeks to complete a house purchase. However, it can take significantly longer and often up to six months. This reflects the range of conveyancing and mortgage processes needed, which are time consuming and administratively cumbersome. Artificial intelligence (AI) and machine learning (ML) algorithms have the potential to automate and streamline these processes, which could make the overall home buying process faster, easier, cheaper, and less stressful. Speeding up the process could also protect buyers and sellers against time wasters, gazumping, and gazundering.

AI and conveyancing: Reducing the paperwork burden

One of the most time-consuming parts of buying a home is the conveyancing. This process covers all the legal requirements of buying a home. Among other things, this includes the property, local authority, and environmental searches, home buyers surveys, drafting and exchanging contracts, as well as dealing with all enquires. These tasks are all administratively heavy and typically take four to twelve weeks to complete.

Generative AI tools can help to reduce the burden. For example, it could be used to produce legal documents and review property contracts to find any issues or missing information. It could summarise lengthy email chains, making it easier for buyers, sellers, and solicitors to stay up to date. AI can also predict and flag any potential upcoming delays in the process. Shifting this administrative burden onto an AI tool can free up solicitors to use their time on more high-value activities.

Machine learning for streamlined mortgage finance

Mortgage finance generally takes between two to six weeks, worked through alongside the conveyancing process to get a full mortgage offer. During the application process the lender will review borrower credentials with the type of property and loan required in mind. These include their employment status, income and affordability, deposit size, and credit rating. This involves evaluating physical documents such as payslips, bank statements, and a passport or driving license.

Some processes are already automated, but ML models, which extract patterns and relationships from data, can further improve the efficiency and timeliness of the process. An ML model could process personalised mortgage documents and make a tailored decision for the lender, adapting its behaviour based on its findings. This can all be done virtually, creating increased efficiency and faster decision-making. One provider which has fully automated the process delivered a fully underwritten offer in 15 minutes.

AI chatbots in mortgage applications and approval speed

Generative AI tools like ChatGPT can also be incorporated into a lender’s mortgage origination platform as an internal chatbot tool for instantly answering complex queries. AI-driven chatbots can then handle inquiries from customers, guide them through the application process, and offer personalised advice 24/7.

Over the past year, there were around 633,000 mortgage approvals, according to the Bank of England. Assuming the average mortgage approval time taken was 28 days, this amounts to over 17 million days spent processing mortgage applications. Even reducing this time by half would save 8.5 million working days.

Of course, the use of AI is not limited to the home buying process. It could even be used to find the perfect home. Searching for the right home can take months and involves assessing a broad number of specific criteria. AI can be used to speed up the process of trawling online property portals. Buyers can search for their ideal home using conversational sentences like “3-bed detached house with a garden near a train station” rather than using manual search filters. The information is then quickly retrieved, and AI can undertake any due diligence checks such as EPC ratings. Further, the use of AI could speed up other aspects and processes within the residential market. For instance, ML models also have the potential to significantly speed up residential planning applications.

Risks of AI in home buying

There are risks from deploying AI in the process. For example, AI models can amplify bias or produce 'hallucinations', where an AI presents false information as fact. Because AI systems learn from their training data, any present biases will influence their decision-making. Without fixing the underlying biases, AI can perpetuate inequality, potentially leading to mortgage approvals being unfairly declined. If AI becomes more widely integrated into the mortgage approval and conveyancing process, then mitigating bias will be a priority.

Ultimately, using AI to speed up the home buying process has the potential to significantly boost efficiency by processing information at a far quicker rate. If widely adopted, this could create a more dynamic and fast-moving market, saving buyers, lenders, and solicitors time and stress. However, it is integral that efficiency is not traded for accuracy and fairness, and that lenders or lawyers using AI can justify their decisions in a transparent way.