One Big Idea

Most banking AI is deployed where the win is easiest to see — not where the money is.

A coding assistant that saves two hours a day is visible — you can put it on a slide. A fraud model that quietly stops a loss, or a pricing engine earning a few more basis points of margin, is real but hard to see — so it gets less of the money. This is the Visibility Trap.

The prize is not small: McKinsey estimates generative AI could add two hundred to three hundred forty billion dollars a year to global banking — nine to fifteen percent of operating profit. But a prize on a slide is not a number on the P&L. The job is not to spend more on AI — it is to spend it where the return lands.

The Insight

By some estimates banks put more than half of their applied-AI budget into back-office efficiency — yet by some estimates more than seventy percent of the audited return comes from elsewhere: avoiding losses and growing revenue.

The reason is the productivity-not-profit leak. Saved time is not profit — it is freed capacity, and capacity becomes money only if the bank redeploys people or reduces headcount. A measured twenty-percent efficiency gain can move return on equity by zero. Efficiency attracts the budget because it is safe and measurable — but measurability is not value.

What you can count is not always what counts. Efficiency wins the budget because it is easy to see; the return lives where it is hard to.

Framework of the Week · The Visibility Trap

A two-by-two: measurement visibility × audited P&L value → four quadrants. Move the money diagonally — out of the trap, into the vault.

  • Visibility Trap — high visibility, low value: where most of the budget goes.
  • Gold Vault — low visibility, high value: where most of the return lives, under-funded.
  • Proven Win — high on both: fund first.
  • Quiet Drain — low on both.

Map every use case to one of 3 value zones: Zone 1 · Run-the-Bank (cost-out) commands the budget but leaks value; Zone 2 · Risk-Intelligence (loss-avoidance) is the highest risk-adjusted return and hits the ledger immediately; Zone 3 · Grow-the-Bank (revenue) has the highest ceiling and the heaviest governance load.

When the budget sits in Zone One, leaders should re-allocate toward the loss-avoidance premium of Zone Two — and pursue Zone Three carefully, with the governance built in.

The full framework, with the quadrants and the value zones drawn out, lives in the Frameworks library.

Use Case · Where the return actually lives

The heaviest technology investor in the industry reportedly spends around two billion dollars a year on AI and recovers about the same — but roughly one and a half billion of that return reportedly comes from fraud, trading, and operational-risk efficiencies (Zone 2). A leading Asian bank reportedly booked around one billion Singapore dollars of value, measured against control groups, driven by personalized engagement (Zone 3). And a US student-loan lender, Earnest, paid a $2.5M settlement over an AI underwriting model that produced disparate outcomes — revenue AI without governance is not value, it is liability.

Risk Note

Revenue AI carries the heaviest governance load — the moment AI touches who gets a loan, fair-lending law is watching. SR 26-2 is most relevant to banks above thirty billion dollars in assets and scales by risk, and it places generative and agentic AI outside its formal scope while still expecting them to be governed — the burden moves to the enterprise. The macro signal: Gartner projects more than forty percent of agentic-AI projects will be cancelled by the end of twenty twenty-seven — most for failing exactly this value test.

Latest Video

This week's video — Most Banking AI Is Deployed Where the Money Isn't — walks the Visibility Trap, the 3 Value Zones, and the Capital Reallocation Loop: tag → trace → re-allocate → guard.

Watch: youtu.be/L37oZW3dY5g

The free five-page playbook in the Frameworks library turns it into a worksheet for your own AI portfolio.

Reply and tell me which zone your AI budget really sits in — most leaders are surprised. I read every response. Forward this to a banking executive whose AI spend is stuck in Zone One.

Was this forwarded to you? Subscribe to The AI Architect Letter — free, every Saturday.

Minh Tran · AI Business Architect · LinkedIn · Workshops & advisory: aibusinessarchitect.ai