Smaller AI models help Bank of America, Capital One manage costs

Major lenders like Bank of America and Capital One are shifting their AI strategy, moving away from massive, expensive 'frontier' models toward smaller, specialized ones. For retailers and operators, this is a signal that the financing tools you use will soon become more efficient and personalized. These smaller models are cheaper to run and faster to deploy, meaning lenders can process applications and manage risk without the massive overhead of general-purpose AI. What this means for your business is a more stable tech environment. When lenders lower their operational costs through tech efficiency, they are better positioned to maintain competitive rates and consistent service levels for your customers. Capital One is specifically focusing on using these models to improve how they interact with users and handle data. Expect to see financing interfaces become faster and smarter as these banks integrate 'Small Language Models' into their approval engines and customer service portals. Rather than one giant AI trying to do everything, lenders are building 'ensembles' of small tools designed for specific tasks like credit scoring or fraud detection.

Source: Banking Dive

← Return to the library· Submit a correction