Credit unions have strong relationships with their members, but Artificial Intelligence represents an extraordinary opportunity to deepen those ties, offer new services at just the right time, and provide even greater convenience to further cement member relationships. AI’s ability to increase a credit union’s engagement with members begins with producing enhanced insights utilizing data that represents member behaviors, patterns, and financial history. Current efforts focus on making members more aware of credit union services available to them by flagging those services at just the right moment. Artificial Intelligence recognizes when a specific service is particularly relevant which then triggers an offer that addresses that specific need for personal credit, savings, bonds, insurance, or even a home mortgage.
In call center operations, AI embedded in existing solutions such as Salesforce recognize when a caller is in distress, in a highly emotional state, or is becoming frustrated during the call which enables intervention. That intervention might transfer the caller from IVR to a live agent or from a live agent to a specialized desk. For example, one financial institution Mercator interviewed is working to train its AI to recognize behaviors associated with elder abuse. AI’s ability to categorize content can also be used to evaluate phone recordings, emails and letters to identify and rank common issues observed in these various channels.
Few credit unions recognize the many Artificial Intelligence solutions are generally available today, either built into software solutions that are already deployed such as Salesforce or available in versions that are pre-integrated into existing solutions. For example, a company called Faraday is already integrated into Oracle and Cogito is pre-integrated and available in the Salesforce AppExchange.
So the implementation challenge shouldn’t begin with concerns for technological competence in AI, the first step is to determine where your institution could benefit from the application of AI and then look for that solution in the market. To help with that first step here is a list of areas where AI can improve operational efficiency and lower risk for credit unions:
- Member Onboarding
- Detecting Account Takeover Attempts
- Mobile and Web Personalization
- Recommendation Management
- Loan Application and Risk Decisions
- Call Center Operations
- Chargeback / Dispute Management Regulatory Compliance
- Video / Audio Compliance Monitoring
- Credit Scoring
- Cash Forecasting
- Collections / Past Due Analysis
Many of the solutions listed go far beyond statistical analysis by utilizing AI in various ways, for example: to extract needed information from existing sources to prefill forms, to validate the data entered by a member, or to automate fraud detection. Note that every AI model is driven by statistics and so every decision made by an AI model is accompanied by a level of confidence in that decision. As a result, any process automated with AI can easily decide when it should escalate a transaction for human intervention. If the AI tool fails to meet the confidence level set by the credit union it can easily be identified and processed manually.
A great example of this sort of AI implementation was introduced earlier this year when CO-OP Financial Services announced the launch of COOPER, an advanced, AI-driven platform for fighting fraud at credit unions. Fraud is a particularly sensitive concern for credit unions and COOPER will arm human fraud analysts with advanced pattern detection to find and prevent fraudulent transactions.
Most everyone is now aware that AI can recognize and respond to text messages and the spoken word. Many of us use Google Assistant and Siri to get answers to our questions. So the question is not if such AI-based solutions will be used by credit unions, but when. These tools will be used to assist credit union employees as they interact with members. They’ll help new employees recognize how to service members while also remaining within the regulatory and credit union guidelines. They will assist managers wrestling with complicated issues that require a comprehensive understanding of regulations and credit union policies and systems. They are already being widely used to reduce risk and loss.
Recent research by the National Institute for Standards (NIST) in conjunction with three different universities indicates that when a professional is teamed with either a human expert or an AI expert that professional’s results increases substantially when interacting with an automated assistant compared to interacting with a human expert.[i] The increased performance is likely due to the fact the trained professional does not expect the machine to be correct but recognizes that the machine may be correct and it has no ulterior motive or personal judgment. Whatever the reason, it clearly points to the benefits of utilizing Artificial Intelligence in an expert system applied in an assistive role. The not so distant future will almost certainly have AI systems interacting with credit union employees across a wide range of operations, making decisions independently and offering advice to guide employees to remain in alignment with credit union guidelines and regulations.
Learn more about the transformative power of Artificial Intelligence and machine learning by downloading the latest whitepaper by Mercator and CO-OP Financial Services, “Accelerating Growth Through AI and Machine Learning”.
[i] https://www.nist.gov/news-events/news/2018/05/nist-study-shows-face-recognition-experts-perform-better-ai-partner
