Wall Street's AI Revolution: Banks Plan for Fewer People

It is December 2025. The era of experimenting with AI in isolated sandboxes is over on Wall Street. During a recent Goldman Sachs conference in New York, the message was loud and clear: AI is operational, productivity is skyrocketing, and yes, that means banks will be doing the same (or more) work with fewer people in the future.

To the outside world, it might still look like science fiction, but within the walls of giants like JPMorgan Chase, Wells Fargo, and Citigroup, the transformation is in full swing. Generative AI is no longer a gadget for the IT department; it has become a fundamental part of business strategy.

The Numbers Don't Lie: An Unprecedented Productivity Leap

Marianne Lake, CEO of Consumer & Community Banking at JPMorgan, shared figures that would make any CFO salivate. Productivity in departments that have fully embraced AI has risen to about 6%. That might sound modest, but it's a doubling compared to the 3% pre-AI. And this is just the beginning.

Lake predicts that certain operational roles will eventually see productivity gains of 40% to 50% as AI becomes more deeply integrated into daily routines. These are no longer marginal improvements; this is a fundamental reshuffling of how much human labor is needed to run a bank.

Software Development in the Fast Lane

At Citigroup, incoming CFO Gonzalo Luchetti sees similar trends. In software development, a productivity improvement of 9% has already been measured. This mirrors a broader trend in the tech sector where "AI Copilots" are taking over code writing, testing, and documentation.

This isn't about replacing programmers, but about eliminating the "boring" work. Generating code, writing unit tests, refactoring legacy code; these are tasks that AI can do faster and often more accurately, as long as an expert is supervising.

The Hard Reality: "Fewer People"

Here is where the story gets serious. If you can produce more with the same team, you have two options: either you grow explosively, or you shrink your workforce. Banks seem to be choosing a combination, but the emphasis is shifting towards efficiency.

Charlie Scharf, CEO of Wells Fargo, didn't mince words. Although the bank hasn't laid off people directly because of AI yet, he noted that they are "getting a lot more done." The bank's internal budgets already point to a smaller workforce by 2026. Scharf also warned of higher severance costs, a clear signal that preparations are being made for future adjustments.

Where AI Has the Most Impact

The biggest gains aren't being made in the flashy trading algorithms that Hollywood movies are about, but in the boring, repetitive processes. It involves work that relies heavily on documents, follows rules, and is predictable.

• Operations: Drafting responses, summarizing files, and resolving exceptions faster.
• Customer Service: Better self-service options resulting in fewer calls reaching agents, and real-time support for agents when they do speak to a customer.
• Sales Support & Onboarding: Extracting data from documents, filling out forms, and speeding up the setup for new clients.
• Regulatory Reporting: Compiling reports and evidence under strict control.

Goldman Sachs' internal "OneGS 3.0" program focuses specifically on these areas: improving sales processes and streamlining process-intensive functions like lending and vendor management.

Governance: The Brake That Builds Trust

Why aren't we seeing full automation yet? The answer is simple: control. Banks operate in one of the most heavily regulated environments in the world. An AI that hallucinates and approves a loan that shouldn't be approved isn't an "oops" moment, but a potential billion-dollar problem.

This is crucial. US regulators (such as the Federal Reserve) require models to be explainable and traceable. "Black box" AI is a no-go in the financial world.

The Future: Phase 2 and Beyond

We are now in what experts call "Phase 1": stable headcount, higher output. Everyone gets a "superpower" in the form of AI tools. But "Phase 2" is coming.

Phase 2 begins when the gains are so consistent that they start influencing workforce planning. This will likely happen through natural attrition (people leaving are not replaced), role changes, and in some cases, targeted cuts.

McKinsey estimates that generative AI could add between $200 billion and $340 billion in annual value to the banking sector. That's a pie too big to ignore. The question is no longer if AI delivers results, but how fast banks can make these gains structural without ruining their risk profile.

💡 Conclusion for the Professional

Do you work in the financial sector? Then the message is clear:

• ➜ Embrace the tools: Be the one steering the AI, not the one being replaced by it.
• ➜ Focus on complexity: Routine work is disappearing. Your value lies in managing exceptions, relationships, and complex decision-making.
• ➜ Understand Governance: Knowledge of how AI is regulated is becoming just as important as knowledge of the AI itself.

Ready to upgrade your workflow?

Join thousands of power users who trust AI Workspace to organize their prompts and conversations securely.

Install for Free