Bank AI Costs Could Reshape Fees and Savings Yields
Retail banking customers and savers may see changes in fees, deposit rates, and access as banks control AI agent spending.
Updated on Nov. 10, 2026 in Banking

Live Poll
Would you let an AI agent make a financial decision for you?
McKinsey reports that banks worldwide are working to curb soaring costs for AI agents as they become more selective about where they deploy agentic AI. Token spending could reach 20 percent of total bank labor costs, prompting some banks to shift oversight to new AI FinOps teams.
Why it matters
Rising AI usage costs can flow through to banks' operating expenses and, in turn, to the banking products and services consumers rely on. Banks are now weighing whether AI-driven revenue growth justifies the spending, and AI agents could also shift deposits into higher-yielding accounts, affecting how savers earn on their cash.
McKinsey estimates token spending could reach 20 percent of total bank labor costs, and 72 percent of surveyed AI users said they would allow a banking agent to complete one financial task. Meta's Muse agent logged 5 million US downloads in its first 22 days, a pace the report contrasts with the 5 to 10 years mobile internet took to gain broad adoption.
The players
McKinsey
A global consulting firm whose research on banking and AI shapes how financial institutions budget for technology and customer services.
Meta
The parent company of Facebook and Instagram, whose Muse agent illustrates how quickly consumers may adopt AI tools that could handle financial tasks.
The details
Banks are routing AI requests to different models based on query complexity as an initial cost control, but that approach breaks down when AI agents communicate with one another, because tokens accumulate quickly across those exchanges. Some banks have responded by moving AI cost oversight to new AI FinOps teams, while others are testing whether AI-driven revenue growth can justify the higher spending. Most banks currently block external AI agents from authorizing transactions within their systems, and surveyed AI users reported a willingness to let agents approve only limited amounts, with one cited cap at $500.
Timeline
Meta's Muse agent recorded 5 million US downloads within 22 days of its launch.
McKinsey expects AI agents to become widely used within three years.
Money Landscape
The rapid early uptake of Meta's Muse agent, compared with the 5 to 10 years mobile internet took to gain broad adoption, suggests banks face a shorter window to adapt their cost structures and customer offerings. McKinsey's three-year expectation for widespread AI agent use places this shift within the current planning cycle for retail banks.
Consumers who already use AI to compare banking services may see more tools that can complete routine transactions on their behalf, and the terms attached to those tools could affect fees and account access. Households should review any bank or fintech permissions that allow external apps to initiate payments before granting them.
The takeaway
AI agents could reshape how banks earn and spend, and that may eventually affect deposit rates and fees. Savers should check whether their bank pays competitive yields on deposits and periodically review which third-party apps hold permission to move money from their accounts.
Further reading
For broader context on how these shifts touch everyday accounts, explore our Banking coverage.
Live Poll
Would you let an AI agent make a financial decision for you?


