AI agents affect bank profits by removing customer inertia and moving money into higher-yield accounts. McKinsey warns that autonomous finance bots could cut global bank profits by up to $170 billion, or 9%, as they optimize deposits and push rate competition. Early adopters can still win by reshaping products, costs, and customer value. Banks face a simple but serious shift. Smart bots can compare rates, move cash, and lock better deals for customers, without friction. McKinsey calls these tools “agentic AI.” They will change how deposits flow, how loyalty works, and how banks earn. Today, checking accounts hold trillions in low interest cash. If AI helps people move even a small slice to top-rate accounts, profits fall. The change is already visible. Big banks are reducing hiring as they bring automation into core workflows. At the same time, most firms are not yet ready to use AI at scale. The gap between those who move fast and those who stall will define competition. The next two years will set the curve for deposit pricing, product design, and customer trust.

How AI agents affect bank profits: the $170B wake-up call

McKinsey’s new analysis estimates a profit hit of up to $170 billion if banks do not adapt to autonomous tools that manage money for customers. The logic is clear. Many people leave funds in checking accounts with near-zero interest. Agentic AI will remove that inertia and move those funds into higher-yield accounts. Here are the core numbers:

Consumers hold about $23 trillion in checking deposits out of roughly $70 trillion in total global deposits.

If just 5% to 10% of those low-yield balances move to top-market rates, it could equal 20% of the industry’s global profits.

McKinsey estimates a potential 9% drop in profits across banks worldwide.

The threat hits the heart of net interest income. When bots nudge or automatically sweep cash into better-yielding products, deposit costs rise, and spreads shrink. Banks may see early cost savings from AI in operations, but competition will pass most of that value to customers.

What agentic AI does for consumers

It removes friction

For decades, banks relied on customer inertia. People rarely changed accounts or hunted for top rates. Agentic AI replaces that habit with action. A bot can suggest a move in seconds and complete it on behalf of the customer.

It improves choices

Bots can compare rates, fees, bonus offers, and safety. They can personalize based on goals, risk, and time horizon. They can time moves around payroll and bills to avoid overdrafts. They can ladder savings or shift funds between banks.

It runs 24/7

Automated agents do not sleep. They track changes in rates, promotions, and market conditions in real time. They can re-balance balances with no extra effort from the customer.

Where profits compress: deposits, fees, and loyalty

Deposit costs go up

When agents move money to better rates, banks must pay more to keep balances. This raises the deposit “beta,” which is the share of rate increases that banks pass to depositors. Higher beta pushes net interest margins down.

Fee income faces pressure

If bots reduce overdrafts, inactivity fees, or out-of-network usage, some fee streams shrink. Better decisions by customers mean less revenue from friction.

Loyalty shifts from brand to value

If a bot shops for the best offer each month, loyalty based on habit fades. Trust still matters, but it shifts toward transparency, rate fairness, and seamless experience. Banks need to earn loyalty with proactive value, not passive inertia.

Playbook: Defend deposits and grow value

Make deposit products work on autopilot

Design accounts that do the smart moves for the customer before outside bots do:

Auto-sweep extra cash from checking to high-yield savings daily or weekly.

Offer tiered rates that improve with balance or engagement.

Provide “set-and-forget” goals like vacation, emergency, and tax buckets with fair rates.

Let customers lock bonus rates for a period in exchange for low-friction commitments.

Be the agent, not just the account

Build or partner to deliver your own agentic experience:

Proactive alerts that say “You could earn $X more by moving $Y today” with one-tap action.

Smart bill timing to avoid overdraft and interest charges.

Rate-matching prompts when a trusted competitor beats your current yield.

Cross-bank optimization if regulations and partnerships allow it.

Reward long-term relationships

Aim for durable ties that beat short-term rate chasing:

Bundle mortgage or card benefits with savings rates.

Offer relationship points that boost yields or cut fees.

Give clear, simple rewards that people understand at a glance.

Use pricing that adapts fast

Real-time pricing engines can match market moves quickly:

Protect core balances with targeted rate boosts.

Focus promotional spend on customers at risk of leaving.

Avoid blanket increases that erode margin across the board.

Build the right tech, data, and risk controls

Clean data and real-time systems

Agentic use cases need fast, accurate data:

Unified profiles for each customer across products.

Event-driven systems that react to changes in minutes, not days.

APIs that let pricing, risk, and servicing talk to each other.

Model oversight and safety

Banks must meet high standards as automation grows:

Strong model risk management for LLMs and decision engines.

Clear guardrails on what agents can do without consent.

Human review for sensitive actions and large transfers.

Transparent logs so regulators can audit decisions.

Privacy and consent deserve center stage

Customers should know how agents act and what they see:

Simple consent flows with plain language and examples.

Easy opt-out and controls for automation levels.

Default settings that protect customer interests.

Workforce, culture, and metrics that matter

Reskill teams for AI-first operations

Banks are already reshaping hiring and roles as they deploy AI. Focus talent on:

Data engineering, model testing, and prompt design.

Product managers who build automation into everyday banking.

Risk, compliance, and audit teams fluent in AI tooling.

Align incentives with customer outcomes

Pay and goals should reflect long-term value, not only quarterly spreads:

Reward net promoter gains and retention of primary relationships.

Track lifetime value, not just short-term deposit margins.

Measure savings delivered to customers as a trust signal.

Key metrics to watch

Deposit beta: How fast do you pass on rate changes?

Churn risk: Which customers are likely to move funds soon?

Product mix: Are balances shifting to high-yield or locked terms?

Automation impact: What share of service and ops is AI-powered?

Complaint trends: Are bots causing errors or confusion?

Scenarios: from mild shift to profit shock

Scenario 1: Mild shift

Agents move a small slice of balances to higher rates. Banks see modest margin compression. Early adopters offset pressure with better pricing tools, retention programs, and cross-sell. Profit impact is manageable.

Scenario 2: Reset

A few major banks launch strong agent features. They market rate fairness and automatic optimization. Competitors match. Deposit beta rises across the market. Net interest margins compress. Fee income drops in some areas. Profits fall, but leaders win share and grow long-term value.

Scenario 3: Profit shock

Popular consumer agents make switching effortless. 5% to 10% of low-yield balances move into top rates within a year. The industry sees a sharp hit close to McKinsey’s $170 billion estimate. Only banks with strong automation, pricing agility, and trusted products protect margin and keep customers.

How to stand out when bots do the shopping

Design with “bot empathy”

Assume an agent will test your pricing, speed, and clarity every day:

Publish clear rates and rules. Avoid hidden catches.

Minimize fees that bots can route around.

Offer one-tap account moves and simple confirmations.

Make value obvious

Bots love numbers. People love certainty. Blend both:

Show “annual dollars earned” on savings pages.

Forecast earnings if the customer moves more funds.

Share monthly summaries of actions that saved money.

Earn trust by default

Customers will grant automation to banks they trust:

Give safety nets: undo windows, alerts, and daily limits.

Explain actions in short, clear language.

Offer human help on demand for complex issues.

Cost savings are real, but most pass to customers

McKinsey expects banks to save 15% to 20% in operating costs as AI scales. This includes faster service, fewer errors, and leaner back-office work. But as more banks adopt these tools, competition will push those savings into better pricing and customer benefit. The early window to capture extra profit will close. Move fast to modernize, then shift focus to retention and revenue quality.

Case signals from industry leaders

Some major banks are already reshaping their approach. They report strong earnings, but they also cap hiring and retool roles as automation expands. Leadership teams point to AI across client service, employee workflows, and back-office tasks. The message is clear: reduce structural cost, improve speed, and prepare for deposit competition driven by smart agents.

Risk, regulation, and responsibility

Prevent harm before it happens

Automated moves can create mistakes if not designed well:

Protect against overdrafts during transfers.

Check for minimum balance rules to avoid fees.

Validate identity and permissions for each action.

Supervise third-party and cross-bank flows

If agents move money between institutions, banks need strong vendor risk controls and clear disclosures. Partner only with firms that meet security and compliance standards. Keep audit trails for every automated decision.

Be fair and inclusive

AI must not create unfair outcomes. Test models for bias. Offer alternatives for customers who do not want automation. Provide education and simple guides. Make better money outcomes standard for everyone, not just for power users.

Practical next steps for the next 180 days

Launch or improve auto-sweep from checking to savings with clear guardrails.

Build a rate-matching playbook for at-risk segments.

Stand up a small “deposit defense” war room with pricing, data, and product leads.

Pilot a simple, safe agent that recommends one high-value action per month.

Create transparent consent and safety controls for automated moves.

Prepare model risk documentation for all agentic features.

Train front-line staff to explain AI actions in plain language.

The long game: from margin to relationship value

The goal is not to protect every basis point of today’s margin. The goal is to keep the primary relationship, grow customer lifetime value, and earn trust in a world where bots do the heavy lifting. If your bank saves customers money before someone else does, you build loyalty that lasts. If you resist, smart agents will find higher value elsewhere.

Bottom line

The next phase of digital banking will be driven by automation that acts on behalf of customers. Leaders will redesign products, pricing, and platforms for a world where decisions happen in seconds. As AI agents affect bank profits, the banks that win will move first, prove value, and make fair outcomes automatic.

(Source: https://www.perplexity.ai/page/mckinsey-warns-banks-could-los-10SnUuySTf6YyVauGZq6Cg)

For more news: Click Here

FAQ