A recent Financial Brand article on AI versus Automation caught my eye. Not just because the topic is relevant and the content excellent, but because the article was written by one of FNBB’s own customers, Corey LeBlanc of Locality Bank, headquartered in Ft. Lauderdale Florida. The focus of the article is how FIs should differentiate when AI versus Automation should be used to tackle a business problem. Corey points out that getting these mixed up is not only inefficient, it can have disastrous consequences. Let’s create some boundary definitions: Automation – a tool or service that operates with pre-defined rules, executes with precision and never deviates. Automation works well for scheduled tasks (think out of office emails, scheduled bill payments or monitoring systems for compliance adherence …). However, Automation struggles with tasks that require context, understanding or adaptive behavior. For example, an automated compliance monitor may flag a specific transaction but not recognize a pattern of multiple transactions across multiple payment systems. AI – a tool or service that continually learns and adjusts. An AI chatbot doesn’t just look for a keyword and return pre-determined text, it seeks to understand sentiment, adjusts to reflect previous learning and offers bespoke results. But AI’s strength comes from the volume and quality of data it has at its disposal. Without proper training and ongoing monitoring, AI used for compliance monitoring can miss events that should be detected. Interestingly, Automation can cover some of the potential “blind spots” of AI but that would only work if you were deploying both to create a systemic approach to compliance management. LeBlanc lays out some banking tasks that are well suited for either Automation or AI. They include: Automation: Routine data entry Basic customer service inquiries Scheduled report generation AI Fraud detection Personalized financial advice Risk assessment There are many AI powered systems on the market targeting financial services. Some of these are focused on a single outcome, such as compliance monitoring. Others are general use large language models, such as ChatGPT, Gemini, CoPilot or services that are based on the underlying technology. It may be difficult for a bank to decide whether they need a specific product designed to achieve a specific outcome or a more general-purpose tool that could be trained to achieve the same or similar outcome. When deciding where to invest in AI or Automation, LeBlanc suggested you consider these factors: Process Complexity: Does the task require adaptability and decision-making (AI) or is it routine and rule-based (Automation)? Data Availability: AI requires high-quality, abundant data to deliver meaningful results, regardless of whether the outcome requires a large language model or a small language model. Return on Investment: Will the benefits of AI significantly exceed those of automation to justify the higher cost? You can actually put a pencil to this and create an expected cost to implement and maintain over a period of time. Scalability: Can the chosen solution grow with your institution’s needs? Hard to quantify but you must make some type of estimation of the longevity of a potential solution you are considering. It is likely that many of the companies actively selling AI solutions right now will not be around in 5 years. Regulatory Compliance: Ensure the technology aligns with legal and industry standards. More importantly, as regulations change, especially as it relates to FIs use of AI, how responsively adaptive will your vendor be? There are numerous areas within the bank where AI and/or Automation can create efficiencies and positively impact the gross margin. One resource is a previous FNBB article that highlighted AI’s potential usage at a bank. You can access that article here. Being diligent about the projects that need Automation and which need AI should get baked into your analysis of each scenario. Using AI where Automation is the right tool or vice versa is more than just inefficient. Failure to get compliance right or making errors on any of the other areas of a bank that are monitored by Auditors and Examiners can result in devastating penalties. Community banks like Locality are central to the health of their communities. Forward thinking bankers like Corey LeBlanc make community banks strong and will remain relevant for their customers for many decades to come. Be like Corey: learn about AI and Automation and become the community bank of the future. Resources https://thefinancialbrand.com/news/data-analytics-banking/artificial-intelligence-banking/banks-must-get-use-of-ai-and-automation-right-or-waste-money-183314/