Goldman had 600 traders in 2000. Today: single digits. JPMorgan is spending $19.8 billion on tech. This is the most comprehensive banking role analysis in my 119-industry series.

Goldman Sachs had 600 equity traders on their floor in 2000. Today? Single digits.

And yet Goldman deployed its AI assistant to all 46,000 employees, partnered with Anthropic to build autonomous AI agents for accounting and compliance, and rolled out an autonomous coding tool across 12,000 engineers — reporting 3-4x productivity gains. JPMorgan is spending $19.8 billion on technology in 2026, with $1.2 billion earmarked specifically for AI. Bank of America’s virtual assistant Erica has handled over 3.2 billion client interactions.

Meanwhile, Citigroup is cutting approximately 60,000 positions by end of 2026. HSBC is eliminating up to 20,000 roles — 10% of its workforce. Bloomberg Intelligence projects 200,000 banking jobs will disappear globally over the next 3-5 years.

An American Banker survey from March 2026 found that only 3% of bankers say AI has actually led to workforce reductions at their firms.

The disconnect is the story.

I analyzed 75 banking and financial services roles across retail banking, corporate banking, trading, treasury, compliance, private banking, and technology — the most comprehensive evaluation in my 119-industry series. The results reveal an industry that isn’t shrinking, but splitting into two parallel universes: one running on algorithms, one running on human judgment.

The Numbers

Category # of roles % of total Avg replacement rate
Fully automated (>90%) 3 4% 91%
Heavy AI assistance (60-90%) 28 37% 69%
Limited AI assistance (30-60%) 32 43% 37%
Irreplaceable (<30%) 12 16% 16%

Industry-wide AI replacement rate: 47% — rising to approximately 52% when weighted by headcount, because the most automated roles (tellers, clerks, clearing staff) employ the most people.

That makes banking one of the most AI-disrupted white-collar industries on earth — comparable to IT and accounting. But unlike those industries, banking has a unique dynamic: regulation simultaneously accelerates AND decelerates automation.

REPLACE Tier: The Roles Being Absorbed Into Platforms

Bank Tellers — 92% automation

The U.S. Bureau of Labor Statistics projects teller jobs will decline 15% by 2033, eliminating over 50,000 positions. Europe has closed 31% of bank branches over the past decade. Bank of America’s Erica handles over 3.2 billion interactions. Intelligent ATMs process 90%+ of traditional counter transactions.

The remaining tellers aren’t processing deposits. They’re “relationship advisors” helping elderly clients and handling complex exceptions. The job title survives; the job doesn’t.

Clearing & Settlement Specialists — 80-92% automation

Real-time payment networks have eliminated the batch-processing era. The Clearing House’s RTP network processed $481 billion in Q2 2025 alone — triple-digit quarterly growth. FedNow enables millisecond-level settlement. AI handles reconciliation, exception detection, and routing automatically.

Paper-based clearing is essentially extinct — China’s electronic bill adoption exceeds 98%, with AI OCR handling residual paper at 99%+ accuracy. These roles aren’t being automated. They’ve already been automated.

KYC Specialists — 78% automation

Know-Your-Customer is one of the deepest AI penetrations in banking compliance. McKinsey’s 2025 report detailed how agentic AI now handles end-to-end customer onboarding: document verification, identity checks, risk scoring, and ongoing monitoring. Perpetual KYC — continuous automated monitoring replacing periodic reviews — is becoming the industry standard.

Onfido and Jumio verify identities in seconds. Moody’s offers end-to-end KYC/KYB/AML automation. But enhanced due diligence on politically exposed persons and complex corporate structures still requires senior analysts. The junior KYC role is disappearing; the senior investigator role is not.

Credit Approval Officers — 78% automation

AI-powered underwriting is the highest-ROI AI application in banking. Zest AI reports 40% lower defect rates and 5-day shorter approval cycles. Banks using AI underwriting process 3-4x the loan volume with the same headcount. 83% of lenders plan to increase their GenAI budgets in 2026 (Celent).

The shift is from “human approves every loan” to “AI approves routine loans, human handles exceptions.” EU AI Act — enforceable August 2, 2026 — will require explainability for AI credit decisions, creating demand for AI model interpreters, not more manual approvers.

Traders — 72% automation

Goldman’s trading floor transformation is the most cited case study in banking AI. AI-driven systems now execute 60-75% of equity trading volume in major markets. JPMorgan’s LOXM algorithm optimizes execution in milliseconds.

But “trader” is not one job. Execution traders are essentially gone. Market-making traders are heavily automated but still needed for extreme volatility. Macro and event-driven traders — making judgment calls about geopolitical risk and central bank policy — are largely untouched. The floor got smaller. The decisions got bigger.

Collection Managers — 72% automation

AI collection platforms deliver 8x operational acceleration and 70% lower contact costs. Skit AI and similar tools automate outbound calls, predict account recoverability, and optimize contact strategies across voice, SMS, and email. The global debt collection software market hit $4.8 billion in 2025 (projected $11.3 billion by 2033). But bankruptcy proceedings and complex restructuring negotiations still require humans.

AMPLIFY Tier: Fewer People, Higher Impact

Financial Analysts — 65% automation

AI automates data extraction, financial modeling, and report generation. The AI accounting market is growing at 44.6% CAGR (Mordor Intelligence). OCR and machine learning auto-process invoices across formats. The “cognitive analyst” concept — querying financial data through natural language instead of spreadsheets — is already operational.

But BCG notes that average AI ROI in finance sits at ~10%, below the 20%+ target most firms set. A third of finance leaders say AI hasn’t delivered significant value yet. The automation curve is early, with massive upside remaining.

Internal Auditors — 60% automation

MindBridge AI analyzes 100% of financial transactions in seconds. Traditional auditing? 5-10% sampling. This isn’t incremental improvement — it’s a category shift. DataSnipper hit a $1 billion valuation with 6,715% revenue growth, now used by 600,000+ audit and finance professionals across 175 countries.

But someone still signs the audit opinion. Someone still explains findings to the board. The execution layer is automated; the judgment layer is not.

Risk Managers — 50% automation

HSBC’s AML AI processes over 1 billion transactions monthly (Google Cloud partnership) and catches 2-4x more suspicious activity than legacy systems, with 60% fewer false alerts. But Oliver Wyman’s CRO survey found that 74% of Chief Risk Officers now rank technology and cyber risk among their top five concerns — a human judgment call about risks that AI itself creates.

Wealth Managers — 35-45% automation

Robo-advisors are projected to manage $5.9 trillion by 2027 (up from $2.5 trillion in 2022). HSBC’s AI advisor handles 1 million queries daily with 35% higher client satisfaction. Vanguard’s robo-advisor beats benchmarks by 2%.

But the “hybrid advisor” model is the consensus, not full automation. AI cuts meeting prep from 4-6 hours to ~1 hour (Financial Planning 2026). That doesn’t replace the advisor — it lets them spend more time on what clients actually value: trust, judgment, and personal attention.

The Compliance Paradox

Here’s the counterintuitive case that keeps repeating across my series: AI doesn’t reduce compliance needs. It multiplies them.

EU AI Act full enforcement begins August 2, 2026. Every AI model a bank deploys creates new audit trails, explainability requirements, and bias monitoring obligations. Fines reach up to 7% of global turnover — or EUR 35 million, whichever is higher.

Citigroup trained 175,000 employees on AI tools. Those employees generated 6.5 million prompts in a mandatory “Prompting like a Pro” program. Every one of those prompts is a decision point that compliance must be able to explain to a regulator.

In April 2026, Federal Reserve Chair Powell and Treasury Secretary Bessent held a call with major US bank CEOs about an AI cyber threat. The compliance officer’s role isn’t shrinking — it’s expanding into AI governance, model risk, and algorithmic oversight. At 35% automation, it’s one of the safest roles in all of banking.

EMERGE Tier: The Roles AI Created

AI/ML Risk Engineers — 15% automation

This is the role AI created, not replaced. Goldman Sachs deployed Anthropic’s Claude for autonomous accounting and compliance work in February 2026, with Anthropic engineers embedded on-site for six months. Goldman also rolled out Cognition’s Devin — an autonomous coding agent — across 12,000 engineers, reporting 3-4x productivity gains.

Someone has to build, monitor, validate, and audit these systems. JPMorgan’s $19.8 billion tech budget is, in large part, spent hiring these people. Salary range: $150,000-$400,000. Demand growing over 50% in the next five years.

RegTech Engineers — 25% automation

The global regulatory technology market is exploding. Sumsub’s 2026 report confirms AI is revolutionizing AML and compliance — but the tools need engineers who understand both code and regulation. This role barely existed five years ago. Now every major bank has a RegTech team.

Open Banking API Developers — 35% automation

Accenture research shows 76% of banks expect open banking usage to grow 50%+ by 2026. AI coding assistants like GitHub Copilot boost developer productivity by 30-50%, but API architecture design, security, and integration strategy require deep human expertise. This is a growth role, not a shrinking one.

The Private Banker Test

The most AI-proof client-facing role in banking? Private banker — at 25% automation.

Serving ultra-high-net-worth clients ($30M+ assets) is fundamentally about trust built over decades, understanding family dynamics across generations, and navigating complex needs — tax optimization, family governance, philanthropic structures — that require judgment, not algorithms.

Citi Wealth launched Advisor Insights and AskWealth AI platforms in 2025. These tools make private bankers more efficient. They don’t make private bankers unnecessary. When you’re managing $50 million of someone’s family wealth, they want a human across the table.

Right behind it: family office managers at 20%. Multi-generational wealth transfer, family governance, charitable planning — irreducibly human.

The 3% Paradox Explained

Why do only 3% of bankers say AI caused layoffs at their firm?

Three reasons:

First, rebranding. Banks call AI-driven cuts “restructuring” or “efficiency programs.” Citi’s 60,000 cuts include approximately 40,000 from a Banamex IPO spin-off. The AI component is real but blended into broader strategic moves.

Second, simultaneous hiring. Banks are aggressively adding AI engineers, ML specialists, and governance roles. American Banker found a 13% increase in AI staff at banks in the past six months. The net headcount change is less dramatic than the gross movements.

Third, the slow squeeze. Banks aren’t firing tellers en masse. They’re closing branches one by one, not replacing departures, and letting natural attrition do the work. It’s invisible to anyone inside the building.

McKinsey estimates GenAI could add $200-340 billion annually to global banking revenue. BCG’s 2026 report concludes: “AI will reshape more jobs than it replaces.” The industry isn’t in denial — it’s in metamorphosis.

What This Means For You

If you’re in banking:

  1. The bottom 40% of roles are being absorbed into platforms. Tellers, clerks, KYC specialists, standard credit analysts — if your daily work is processing transactions or applying rules, AI already does it faster. The timeline isn’t years. It’s months.

  2. Regulation is the career moat. EU AI Act, Basel III.1, PSD3 — every new regulation creates demand for people who understand both the technology and the rules. Compliance, risk, and AI governance are the growth areas.

  3. The C-suite is safer than ever. CEOs (5% automation), CROs (12%), CFOs (15%) — these roles require judgment, accountability, and human relationships that AI cannot replicate. AI makes senior leaders more leveraged, not less relevant.

  4. The biggest opportunity is building the AI. AI/ML engineers, AI audit specialists, and RegTech developers command $150K-$400K salaries with the lowest replacement risk in the entire industry. These roles didn’t exist a decade ago.

Banking isn’t dying. Banking jobs that can be reduced to algorithms are dying. And the 3% stat? It’s not denial. The transformation is so smooth that most people inside the building can’t feel the floor shifting beneath them.

This is part of my 119-industry AI replacement analysis series, based on the Replace / Amplify / Emerge framework. I’ve analyzed 75 banking roles across retail, corporate, investment, private banking, treasury, compliance, and technology.

Previously: HR, Software/Tech, Finance (overview), FinTech, Sales & Marketing, Accounting/Audit/Tax.

Follow for the next analysis: Legal.

Sources

  • American Banker — Only 3% of Bankers Say AI Caused Layoffs (March 2026 survey): https://www.americanbanker.com/news/bankers-say-ai-is-not-eating-jobs-yet
  • American Banker — Banks Increasing Hiring in Age of AI: https://www.americanbanker.com/news/exclusive-research-banks-are-increasing-hiring-in-age-of-ai
  • Bloomberg Intelligence — 200,000 Banking Jobs Projected to Disappear
  • JPMorgan — $19.8B Tech Budget 2026, $1.2B AI: https://www.prismnews.com/news/jpmorgan-boosts-2026-tech-budget-to-198-billion-with-12-billion-for-ai
  • JPMorgan — $2B AI Cost Savings: https://www.investing.com/analysis/jpmorgan-chase-ai-spending-targets-cost-savings-and-margin-gains-200678084
  • Goldman Sachs — 600 Traders to Single Digits: https://www.technologyreview.com/2017/02/07/154141/as-goldman-embraces-automation-even-the-masters-of-the-universe-are-threatened/
  • Goldman Sachs — AI Assistant Deployed Firmwide (46,000 Employees): https://www.cnbc.com/2025/01/21/goldman-sachs-launches-ai-assistant.html
  • Goldman Sachs — Anthropic Claude for Accounting/Compliance (Feb 2026): https://www.cnbc.com/2026/02/06/anthropic-goldman-sachs-ai-model-accounting.html
  • Goldman Sachs — Cognition Devin Autonomous Coder (12K Engineers, 3-4x Productivity): https://www.cnbc.com/2025/07/11/goldman-sachs-autonomous-coder-pilot-marks-major-ai-milestone.html
  • Citigroup — 60,000 Job Cuts by End of 2026: https://www.btimesonline.com/articles/176602/20260204/citigroup-plans-60-000-job-cuts-by-2026-as-automation-redefines-global-banking-work.htm
  • Citigroup — 175,000 Employees AI Training, 6.5M Prompts: https://fortune.com/2025/10/01/citi-ai-prompt-training-mandate-employees-reskilling-workforce-business/
  • Citigroup — CEO Jane Fraser “Not Graded on Effort”: https://fortune.com/2026/01/14/citigroup-ceo-jane-fraser-job-cuts-ai-bad-old-ways-restructuring/
  • HSBC — 20,000 Job Cuts in AI Overhaul: https://www.gbnews.com/money/hsbc-plans-job-cuts-ai-overhaul
  • HSBC — AML AI 1B+ Transactions Monthly (Google Cloud): https://cloud.google.com/blog/topics/financial-services/how-hsbc-fights-money-launderers-with-artificial-intelligence
  • Bank of America — Erica 3.2B+ Interactions (March 2026): https://newsroom.bankofamerica.com/content/newsroom/press-releases/2026/03/bofa-ai-and-digital-innovations-fuel-30-billion-client-interacti.html
  • BLS — Bank Teller Jobs -15% by 2033 (50,000+ Positions): https://www.bls.gov/ooh/Office-and-Administrative-Support/Tellers.htm
  • Finastra — 98% Financial Institutions Using AI (Only 2% Report Zero): https://www.finastra.com/press-media/ai-tipping-point-reached-just-2-financial-institutions-report-no-ai-use-finds-finastra
  • McKinsey — GenAI Could Add $200-340B Annually to Banking: https://www.mckinsey.com/industries/financial-services/our-insights/global-banking-annual-review
  • McKinsey — Agentic AI in Banking KYC/AML: https://www.mckinsey.com/capabilities/risk-and-resilience/our-insights/how-agentic-ai-can-change-the-way-banks-fight-financial-crime
  • BCG — “AI Will Reshape More Jobs Than It Replaces” (2026): https://www.bcg.com/publications/2026/ai-will-reshape-more-jobs-than-it-replaces
  • EU AI Act — August 2, 2026 Full Enforcement (High-Risk Systems): https://ai-act-service-desk.ec.europa.eu/en/ai-act/timeline/timeline-implementation-eu-ai-act
  • Zest AI — 40% Lower Defect Rates, 5-Day Shorter Approval Cycles: https://www.zest.ai/product/underwriting/
  • Celent — 83% Lenders Plan to Increase GenAI Budgets
  • MindBridge AI — 100% Transaction Analysis: https://www.mindbridge.ai/
  • DataSnipper — $1B Valuation, 6,715% Revenue Growth, 600K+ Users: https://www.prnewswire.com/news-releases/datasnipper-delivers-1-4b-in-productivity-savings-in-2025-as-audit-and-finance-enter-the-ai-era-302661257.html
  • Oliver Wyman — 74% of CROs Rank Tech/Cyber Risk Top 5: https://www.oliverwyman.com/our-expertise/insights/2025/nov/cro-survey-key-risks-strategy-shifts-ai-impact.html
  • Fortune — AI Reshaping Banking, Not Causing Mass Layoffs: https://fortune.com/2025/12/23/ai-reshaping-banking-not-causing-jobs-wipeout/
  • The Clearing House — RTP Network $481B Q2 2025
  • Mordor Intelligence — AI in Accounting Market CAGR 44.6%: https://www.mordorintelligence.com/industry-reports/artificial-intelligence-in-accounting-market
  • EY-IIF — Global Bank Risk Management Survey: https://www.ey.com/en_gl/insights/banking-capital-markets/ey-iif-global-bank-risk-management-survey
  • Financial Planning — AI Cuts Advisor Prep Time from 4-6 Hours to ~1 Hour
  • Accenture — 76% Banks Expect Open Banking Growth 50%+
  • InvestSuite — Robo-Advisors $5.9T AUM by 2027: https://www.investsuite.com/insights/blogs/how-robo-advisors-are-reshaping-private-banking-in-2025
  • Citi Wealth — Advisor Insights and AskWealth AI Platforms
  • Sumsub — AI Revolutionizing AML and Compliance (2026): https://sumsub.com/blog/ai-in-anti-money-laundering-and-compliance/
  • Powell/Bessent — AI Cyber Threat Call with Bank CEOs (April 2026): https://www.cnbc.com/2026/04/10/powell-bessent-us-bank-ceos-anthropic-mythos-ai-cyber.html
  • Global AI in Banking Market — $45.6B in 2026, $143.6B by 2030: https://www.blott.com/reports/ai-use-cases-in-banking