At FinovateFall earlier this month, there was one question brought up in multiple panel conversations and on the networking floor: in the age of AI, who owns the customer?

Historically, banks have had an obvious, structural advantage. They hold the checking account, receive the direct deposit from the paycheck, see transactions, and often serve as the place consumers turn when they need to dispute a payment, borrow for a home, finance a major life event, start investing, etc.

The LLM’s role

However, what happens when the place consumers go first with their financial questions isn’t their bank or even a fintech? With LLMs offering increasingly powerful financial capabilities and reliable bank account connections, a customer’s first stop for financial information may be their preferred LLM.

In 2026, consumers have become comfortable using LLMs for a vast array of tasks. People are using LLMs to research questions, plan vacations, manage work projects, interpret health and fitness data, meal-plan, get parenting and school help, navigate relationships, make purchases, and increasingly analyze financial information.

The retailer’s role

Before taking a deeper look at LLM’s relationship with consumers, let’s examine another stakeholder in this equation, the retailer. Retailers, especially big box retailers like Walmart, Costco, and Target, have always held a more intimate level of information about consumers than banks have. They not only see how much the consumer is spending, but they also see each individual item the consumer is purchasing.

So while banks have valuable data regarding their consumers’ financial lives, and retailers see the exact items in customer’s basket, LLMs get to form a picture of customers through their actual lives. Oftentimes, consumers will consult their preferred LLM about details before large purchases and events such as a move, vacation, wedding, new baby, home renovation, or job change. Because of this, the LLM has the data about the consumer before the resulting transactions even appear on their bank statement.

As an example, a bank may see the consumer spent $2,389 at Costco, while Costco sees a consumer who purchased patio furniture, kids’ clothes, hummus, and carrots. That item-level data gives retailers context banks traditionally haven’t had. Retailers can distinguish grocery spending from clothing or home improvement even when everything happened at the same merchant. To summarize, the bank sees the transaction, the retailer sees the purchase, and the LLM sees the person.

The steakhouse that got me thinking

The LLM introduces another layer to this because it understands the rationale behind the purchase. To offer up a personal example, I recently planned my in-laws’ 50th wedding anniversary celebration in Colorado. Because I am not local to the area, I used ChatGPT to help plan details such as photography, celebration locations, a restaurant, and logistics. Ultimately, after using ChatGPT to compare the options, we chose Twin Owls Steakhouse in Estes Park and it was fabulous.

Weeks later, I was reviewing my finances that are connected to ChatGPT via Plaid. On a monthly basis, ChatGPT offers a written overview of my finances with red flags and suggestions. In this report, ChatGPT identified that my restaurant spending for the previous month was higher than normal and exceeded my monthly stated budget. Unprompted, it went one step further. It recognized the Twin Owls Steakhouse charge as the anniversary dinner I had previously planned using ChatGPT. Instead of simply flagging the category as overspending, the LLM distinguished an intentional, one-time celebration from a potentially concerning spending trend.

My bank knew I spent money at Twin Owls. ChatGPT not only knew why I was there, it knew I had planned to spend the money before I even spent it, and crucially, it helped inform the decision of where I ultimately spent the money.

Banks’ new role

So where does this leave banks? When operating alongside LLMs, banks have three options:

• Build: Banks can create AI experiences that sit within their app or website that develop meaningful customer context that extends beyond answering questions about the customers’ transactions.
• Join: Banks can make bank data, products, and execution available within the existing LLMs and AI ecosystems where customers already go to make decisions.
• Differentiate: Banks can double down on what external LLMs cannot independently provide, such as regulated custody, trusted execution, credit, identity, fraud protection and dispute resolution. These services are hard to replicate in an LLM setting and banks retain a significant advantage in these areas.

For years, banks have invested in creating a 360-degree view of the customer. But the next battle for primacy will be won by the company with the most comprehensive view of the customer’s life first.