HOW LENDERS CAN SAFELY ADOPT A COMPETITIVE AI/ML STRATEGY

Artificial intelligence and machine learning are transforming mortgage lending. Is your organization ready?

Artificial intelligence (AI) and machine learning (ML) technology is progressing at warp speed with new breakthroughs announced almost daily. While AI/ML technology is rapidly advancing, adoption in the highly regulated financial services industry has been slow. Lenders are determining how they can safely adopt an AI/ML strategy while addressing its complex regulatory compliance challenges.

Today’s mortgage lenders primarily use AI/ML to address routine problems like comparing and verifying documents. This just scratches the surface of its capabilities. The next frontier has AI/ML applying probability calculations to increasingly complex, subjective data on human behavior to help lenders make better lending decisions.

As AI/ML becomes more autonomous it raises questions, especially among regulators, about the opportunity for discrimination – intentional and unintentional. To mitigate this risk, regulators are requiring lenders to “show the math” by documenting and reporting not only how AI/ML is being used in credit decisions, but what data sources are being used, which algorithms are being applied, and what systems are in place to monitor and test for systemic bias.

At Black Knight, we have been working with lenders across the country to help them develop strategies to safely expand their use of the latest AI/ML technologies. This technology is different from traditional mortgage software that uses structured, rules-based processing to determine outcomes.

In contrast, AI/ML analyzes vast amounts of data using complex algorithmic logic – almost like a super-human brain.  AI/ML explores many data points to make intelligent decisions and learns and evolves over time to continually improve its performance. Adopting AI/ML must take these differences into account, especially when anticipating future regulatory oversight of the algorithms and decisioning regarding fair lending practices.

Why Adopt An AI/ML Strategy?

AI/ML technology encompasses everything from virtual assistants to task automation and fraud detection. There are many competitive advantages gained from adopting an AI/ML strategy, including:

• Reduced Operational Costs: AI/ML can automate routine tasks in the lending process to increase efficiency and learn to improve performance over time.
• Increased Employee Bandwidth: Digital assistants and self-service tools can support consumer inquiries, reducing demands on staff.
• Improved Customer Experiences: AI/ML can provide real-time, highly personalized customer experiences anytime, anywhere through virtual assistants and the latest digital tools.
• Better Documentation Validation and Management: AI/ML can rapidly validate documents for consistency, completeness, accuracy and correctness.
• Predictive Analytics: AI/ML can analyze data from various sources to provide predictions and insights into market trends and customer behavior.
• Enhanced Risk Assessment: AI/ML can analyze vast amounts of data to help lenders make better risk assessments and underwriting decisions.
• Fraud Detection: AI/ML can detect potential fraud in mortgage applications.
• Sales and Marketing: AI/ML can proactively identify opportunities and track consumers’ actions to target advertising.
• Customer Retention: AI/ML can identify customers that require attention in your portfolio to increase retention.

Start with the Data

When developing an AI/ML adoption strategy, data quality, infrastructure and management are the foundations for success. Because AI/ML models require vast amounts of data to make decisions, learn and evolve, implementing your AI/ML strategy should begin with data and data architecture. Lenders need to build an infrastructure that facilitates the sharing of data with their AI/ML models.

Since data is the currency of AI/ML, data governance and effective storage strategies are crucial. Large amounts of data about all the objects used to create models, such as images for document processing and voice or video files, can result in complex storage requirements. Stakeholders in information security, legal and compliance demand governance and controls engineered to protect, anonymize and appropriately use client and consumer data.

Design AI for Regulatory “Explainability”

AI/ML has recently been subject to heightened concern by regulators. While regulatory guidance is currently not definitive, model explainability and transparency are focus areas for regulators and other stakeholders. Your AI/ML strategy must consider the regulatory compliance challenges related to AI/ML decisioning models. Existing regulations and auditing processes are currently based on testing known inputs versus expected outcomes. The AI/ML “black box” technology can be more difficult to defend regarding fair lending rules and other laws.

To proactively address future regulatory demands, lenders should take an active role in developing AI algorithms with stronger explanatory capabilities from the inception. Explainability requires more than a description of a model’s features. Data scientists need to clearly identify the source and composition of training data, provide transparency about how models operate, and be able to document and report on measures taken to detect and correct model bias.

Measure and Tune Model Performance

Unlike traditional software, AI/ML is not a set-and-forget technology. A regular review of model performance and end-user feedback is an important part of your strategy. AI/ML managers should conduct regular reviews of the data used for training to ensure that it is sufficiently representative of production data. Managers need to have processes to quickly evaluate the performance of new models in production and ways to take corrective action if needed.

Invest in Stakeholder Education

Managers should also invest time providing stakeholders across the organization with educational materials and regular briefings. Using AI/ML will impact executive decision-makers, compliance, legal, human resources (HR) and audit teams and a successful program will provide them with knowledge of AI/ML-specific risks and issues.

Conclusion

AI/ML can deliver faster insights, greater operational efficiency and enhanced customer experiences across the loan lifecycle. To use AI/ML effectively, your objectives, data, technology and stakeholders must be aligned. Planning today for future regulatory requirements can help you safely and confidently adopt more sophisticated AI/ML technology to maximize your operational performance and thrive in a competitive mortgage market.

Increasingly, lenders are outsourcing AI/ML planning and services to companies like Black Knight to guide them in this important transformation. While AI/ML is unique, with research, sound planning and the right investment, it can provide many significant competitive advantages to take your business to the next level.  To learn more about AI/ML in the mortgage industry and  how you can safely adopt it in your technology strategy, read our white paper here.