As humans, we are always looking for ways to make our lives easier. That’s natural. After all, we want to spend more time doing the things that matter most to us, like interacting with our loved ones and deepening relationships with our customers. This is becoming a reality every day as the power of artificial intelligence (AI) embeds it into every aspect of our lives.
The real estate and mortgage industries are no exception to the AI wave. Buying a home is one of our biggest transactions. It can often be the most stressful. So today’s question is: How can and should modern tools like AI be used to streamline the entire process?
In its early stages, AI primarily enhanced credit risk models, automated underwriting decisions, and more accurately predicted a borrower’s likelihood of default with the help of machine learning algorithms. But then generative AI and conversational platforms like ChatGPT came along and changed the game for all of us. Suddenly, AI is increasingly integrated into more visible components of the mortgage customer journey. Lenders are quickly announcing that they are relying on AI-driven virtual assistants and chatbots to collect and analyze data, manage borrower inquiries, and support back-office staff by presenting information in a human-like conversational format.
Today, jobs such as loan officers, assistants, and processors are being revamped to form a partnership between machine and human that is expected to benefit the borrower. AI is stepping in to enhance that job, especially thanks to the rise of agent AI, systems that can act autonomously without direct human guidance. These capabilities are now enabled by a growing array of cutting-edge innovations working together, including:
- Machine learning algorithms that can analyze historical loan data to distinguish patterns and enhance risk assessment. This allows lenders to reduce human bias when purchasing loans and make credit decisions more quickly.
- Large-scale language models that disentangle and decipher information from loan applications, real estate appraisals, and other text-heavy documents.
- Agent AI workflows that can order credit reports, submit disclosures, and tackle other tedious tasks.
- Computer vision that can read and inspect identification documents, identification photos, and other visual documents.
We are becoming increasingly aware of how these resources are collectively revolutionizing the mortgage process, automating many of the necessary steps in seconds. This could enable more predictive, data-based underwriting decisions that benefit the borrower experience.
But where is the evidence that AI can make a difference in the mortgage space?
Well, you don’t have to look far. According to research from McKinsey, lenders that use AI can reduce the average timeline for loan approval from 37 days to just 14 days. According to a 2023 Fannie Mae report, financial institutions that implemented a comprehensive AI mortgage solution saw a 41% reduction in processing time and a 29% reduction in operating expenses.
We are seeing many applications of AI in the mortgage industry, including:
- Integrate the home buying process: Some homeownership services companies have reimagined the homebuying experience with platforms that allow consumers to search for a home, apply for a loan, and manage their mortgage, all within an AI-powered ecosystem. Many of these platforms can also connect borrowers with bankers to assist with the application.
- Streamline broker workflow: Some mortgage lenders are rolling out new AI-powered tools for brokers, from virtual assistants that handle calls, schedule appointments, and follow up with past clients to analytical tools that review competitors’ loan quotes and highlight opportunities for brokers to offer better deals to win a borrower’s business.
- Strengthen and accelerate evaluation: In addition to using machine learning to calculate loan-eligible home values faster and more accurately, some of today’s appraisal companies are leveraging tools like AI-powered automatic collateral analyzers to advance their capabilities to streamline appraisal reviews and improve accuracy, helping lenders reduce appraisal times and close loans faster.
These are just a few examples of how rapidly AI adoption is accelerating across industries. According to Stratmor Group, 38% of financial institutions will use AI and machine learning in 2024, up from just 15% a year ago. We also found that 48% of lenders deployed robotic process automation (RPA) tools in 2023, compared to 30% in 2020, even before the rapid emergence of agent-based AI solutions.
I am continually amazed by the potential for AI advancements in the mortgage space. It’s amazing that we live in a world where deep learning, computer vision, and large-scale language models can work seamlessly to examine photos of a property and not only determine what rooms and features are present, but also assess condition and quality according to valuation criteria. AI can generate accurate floor plans from property scans, helping consumers understand whether a home is right for them, while improving data fidelity for financing decisions. These advances will help all of us make faster and more informed decisions.
What makes AI especially great is its ability to extract value from large amounts of unstructured real estate data and other complex and large data sets, and distill that data into actionable insights, such as determining the best time to negotiate with a potential buyer or predicting future real estate value trends. We believe this will improve the accuracy, consistency and accessibility of the entire lending process by leveraging a consistent and reliable methodology at scale.
We have been battling 30-45 day closures for a long time. But I agree with experts who predict that AI tools could shorten closing schedules from weeks to even days in the coming years. My gut feeling is that the long-held goal of a seamless, real-time mortgage process and the ability to close loans based on the speed of life rather than the speed of lending may finally be within reach.
But for every voice defending AI as an enabler that frees mortgage professionals from repetitive tasks and administrative chores, there are also passionate voices warning us about AI as a job-killing disruptor. It’s up to us to proactively define how we adjust and evolve our role as humans in mortgage lending in the post-AI world. Otherwise, you run the risk that increased efficiency translates into fewer job opportunities.
There are also serious concerns about trusting AI-generated industry insights and summaries over the original data source material, especially given the potential for omissions, errors, and flawed data caused by hallucinations. As AI-powered mortgage solutions mature, it will be important to ensure human oversight, transparency, and accountability.
Will AI be the enabler we hope it will be, one that strengthens existing professions and positions, or will it emerge as a disruptive force that fundamentally reshapes the role of the industry? The answer will depend on how mortgage professionals and lenders decide to adapt.
Yes, we can both embrace the potential of AI and clearly define the critical areas where human proficiency remains essential. By striking the right balance, the mortgage sector has a compelling opportunity to improve accuracy, fairness, and convenience for borrowers while navigating the inevitable changes ahead.
