Bank of America urges banks to fix data before chasing AI
Before chasing AI, Bank of America wants banks to fix their data first. A lot of the conversations around artificial intelligence tend to circle back to one question: which institution has built the smartest, most capable model?
Bank of America argues the model quality isn’t the point. An AI feature is only as reliable as the data behind it. That reliability determines whether AI can be trusted to support decisions or should be kept out of high-stakes ones entirely.
The hidden cost of bad data
Capital misallocation is the real risk. Bank of America has built its AI strategy to avoid the trap of treating AI as a shortcut to efficiency, an approach that often results in costly, misguided investments. Matthew Davies, Head of Global Payments Solutions at the bank, notes that the pressure every company now feels to adopt AI fast or risk falling behind is itself the real danger.
“The biggest risk and challenge is misinvestment rather than underinvestment,” he says.
Davies traces that misinvestment risk back to fragmented data scattered across ERP systems, treasury platforms, bank portals, and acquired businesses. Because these systems speak different languages, even the most advanced AI layered on top struggles to reconcile the underlying information. Davies emphasizes that without high-quality, standardized data, there’s no real foundation for automation, forecasting, or any AI solution built on top of it.
Tearsheet’s recent reporting on Intuit Credit Karma echoes the same idea. Rather than building standalone AI assistants, the company built them on a shared view of a customer’s financial life, recognizing that better AI starts with better data and context.
Why standardization comes first
For Bank of America, improving data quality delivers value long before any AI enters the picture. Standardized data alone reduces the manual reconciliation, duplicate entries, and reporting errors that consume hours of finance teams’ time every week.
Davies outlined the bank’s order of operations: standardize the data first, automate repetitive tasks next, and only then pursue bigger AI initiatives. This sequence is less flashy than a full-scale AI rollout, but it addresses the messy reality of modern corporate infrastructure. A company cannot simply paste a neural network over a pile of disjointed records and expect clean results.
The bank’s internal tool, EricaAssist, serves as an example of this data-first approach. EricaAssist functions as an employee copilot grounded in years of client history. It does not guess; it relies on a structured dataset that the bank has spent years refining. This ensures that the assistant provides accurate context rather than hallucinations.
CEO Brian Moynihan has added his own caution toward frontier models. His skepticism toward cutting-edge, large-scale AI reflects a broader concern about safety and control. By treating AI as a supportive step in the process rather than the one making high-stakes decisions, the bank aims to maintain oversight.
This philosophy suggests that the next wave of banking innovation will not belong to the companies that claim the most impressive algorithms, but to those that can organize their information reliably. A clean database is a boring asset, but it is also a durable one. Without it, the flashy AI features eventually break under the weight of bad information.
Some financial institutions are already pivoting toward these infrastructure-heavy approaches. [1] Virtual card usage has surged as providers like Mastercard leverage deep data to build safer, more flexible tools for their users.
Investors are watching these trends closely. [2] US companies expanding AI-native Microsoft operations signal a broader shift where stability and data integrity are becoming just as valuable as raw processing power.
