According to PYMNTS.com, Wells Fargo CEO Charlie Scharf, speaking at a Goldman Sachs conference on Tuesday, December 9, revealed the bank’s engineers are now 30% to 35% more efficient at writing code using generative AI tools. Scharf stated that while this hasn’t led to job cuts yet, teams are more productive, and he sees many other areas, like compliance, legal work, call centers, and investment banking, where AI will allow the company to operate differently with fewer workers. He also indicated the bank will likely report higher severance expenses in Q1 as part of ongoing cost-cutting, noting Wells Fargo had over 210,000 employees at the end of September. Meanwhile, PYMNTS Intelligence research found that 75% of bank customers want greater personalization, which AI could help provide. However, in a separate report, 54% of workers surveyed said AI presents a “significant risk of widespread job displacement,” with 38% specifically fearing AI could eliminate their own job.
The Efficiency Play
Here’s the thing: Scharf’s comments are a brutally candid roadmap for the next phase of corporate AI adoption. It starts with a “force multiplier” for highly-paid specialists like engineers, where a 35% efficiency gain is a huge win that doesn’t immediately threaten headcount. But the real target, as he openly admits, is the vast middle and back-office operations—compliance, legal, call centers. These are areas with massive labor costs and, frankly, processes that are often rule-based and document-heavy. Perfect for AI to ingest and streamline. The logic from a CEO’s chair is coldly simple: if a team of 10 can now do the work of 13, you don’t need to hire those extra 3. And over time, as people leave, you might not backfill all of them. That’s how “efficiency” slowly morphs into “reduction.”
Worker Fears vs. CEO Reality
Now, the fascinating tension is in the PYMNTS worker survey. Scharf is basically outlining the workers‘ fear as a corporate strategy. The survey found that people who use AI more are more afraid it will take their job (50% of weekly users vs. 24% of unfamiliar users). That’s counterintuitive, right? You’d think knowing the tool would make you feel safer. But it probably means that when you work with it daily, you see its potential—and its limitations—with terrifying clarity. You understand exactly which parts of your job are automatable. The CEO sees a bottom-line opportunity; the employee sees a looming obsolescence. Both are probably correct from their own vantage point.
The Personalization Paradox
And this is where it gets tricky. The same AI that might displace call-center workers is also, according to the research, the key to winning back 72% of customers through personalized service. So the narrative flips from “AI as cost-cutter” to “AI as customer-retention tool.” But you have to ask: who builds and maintains those sophisticated, trust-building AI interfaces? It likely won’t be the same number of people currently reading scripts in a call center. It’ll be a smaller, more technical team. The promise of “cognitive banking” is real, but its implementation looks a lot like another round of the efficiency play, just dressed up in marketing terms. The jobs that are created will be different, and probably fewer in number, than the ones that are phased out.
The Bottom Line
So what’s the takeaway? Wells Fargo is giving us a corporate case study in real-time. AI adoption isn’t some vague future threat; it’s a present-day productivity tool with a clear path to impacting staffing. The severance expenses Scharf mentions aren’t an accident—they’re a direct result of this strategy. For workers, the survey data suggests your anxiety is rational, especially if you’re already using these tools. The hope, always, is that new roles emerge. But history tells us that transitions are messy and painful. For now, Scharf’s vision and the workers’ fears are two sides of the same coin. One is just printed in the annual report, and the other is felt in the daily grind.
