Wall Street's biggest banks are pouring tens of billions of dollars into artificial intelligence, fundamentally reshaping how they operate from trading floors to back offices. JPMorgan Chase, Goldman Sachs, Citigroup, Wells Fargo, Bank of America, and Morgan Stanley are all racing to embed generative AI (AI systems trained on vast amounts of text to generate human-like responses) across their operations, though their strategies and spending levels vary significantly.

The financial services industry's AI investment reflects a broader belief that generative AI is now table stakes for competitive survival. Yet despite the massive spending, executives continue to face tough questions about whether the investments will actually deliver returns. JPMorgan CEO Jamie Dimon acknowledged this tension, noting that his bank doesn't "uniquely benefit from AI" since everyone is now using it, suggesting that the technology alone won't provide lasting competitive advantage.

Jamie Dimon

How Are Banks Deploying Generative AI Across Their Operations?

Banks are implementing AI tools across remarkably diverse functions, from automating routine tasks to supporting high-level decision-making. The scale of deployment is staggering, with some institutions already reporting thousands of active use cases. Here's how major financial institutions are rolling out these technologies:

• JPMorgan Chase: The nation's largest bank has already deployed nearly 1,000 use cases spanning fraud protection, marketing, note-taking, and portfolio management. The bank rolled out its proprietary generative AI platform to more than 200,000 employees and is tracking developer adoption through dashboards that classify engineers as "light," "heavy," or "non" users of GitHub Copilot, a coding assistant powered by AI.
• Goldman Sachs: The bank is working with Anthropic, an AI safety company, to develop AI agents (autonomous software systems that can perform tasks independently) that automate internal functions including accounting for trades and client onboarding. Goldman allocated $6 billion to technology spending this year, though CEO David Solomon said he wished it were at least $8 billion.
• Citigroup: Nearly 90% of Citi employees are using AI tools, according to CEO Jane Fraser. The bank trained 4,000 employees as "AI stewards" to embed the technology throughout its business lines and has unveiled an AI wealth advisor while scaling the use of AI agents.
• Wells Fargo, Bank of America, and Morgan Stanley: Wells Fargo launched an "AI teammate" for financial advisors designed to increase daily activity. Bank of America has approved more than 300 AI and machine learning cases with at least 34 generative AI use cases fully deployed. Morgan Stanley's DevGen.AI tool saved developers more than 280,000 hours between January and June alone.

What Are the Real Returns on These Massive Investments?

Banks are claiming significant productivity gains, though measuring return on investment remains challenging. JPMorgan's $2 billion AI investment has already matched its cost in savings, according to Dimon, though the bank didn't specify the timeframe or methodology for this calculation. Wells Fargo CEO Charles Scharf previously stated that generative AI tools have made the bank's engineers up to 35% more productive.

Morgan Stanley's experience offers more concrete evidence. The bank's DevGen.AI tool, which helps developers understand outdated code, saved developers more than 280,000 hours between January and June. That's equivalent to 11,666 days of work that developers would have previously spent deciphering legacy systems. Bank of America's virtual assistant, Erica, has been interacted with more than 3.2 billion times since launching in 2018, suggesting sustained user engagement with AI-powered financial tools.

However, the broader question of whether these investments justify their costs remains contested. Analysts continue to press executives on returns and safety concerns, particularly given the scale of spending. Goldman Sachs CEO David Solomon's comment that he "can't afford" to increase technology spending to $8 billion because "I've got to deliver returns" suggests that even major banks are grappling with the tension between innovation investment and shareholder returns.

How Are Banks Approaching AI Strategy Differently?

While all major banks are investing heavily in generative AI, their strategic approaches reveal important differences in how they're thinking about implementation and employee adoption. These variations suggest that there's no single "right way" to deploy AI at scale in financial services.

Citigroup has prioritized employee buy-in above all else, taking what Tim Ryan, the firm's technology chief, described as a mix of "metrics and pride." Rather than measuring granular individual usage like JPMorgan does, Citi measures "the big things" and encourages employees to use the lowest-cost AI model that meets their needs. This approach reflects a belief that sustainable AI adoption requires cultural alignment, not just top-down mandates.

JPMorgan, by contrast, is taking a more data-driven approach to tracking adoption. The bank updated job objectives for its engineers to explicitly require them to "drive excellence" by adopting AI. In February, JPMorgan reorganized its commercial and investment bank to "maximize the impact of AI," with each major business now reporting to Guy Halamish, the newly named chief operating officer of the Commercial and Investment Bank division.

"I believe firms need a mix of metrics and pride," said Tim Ryan, Citigroup's technology chief overseeing the bank's $12 billion tech budget.

Tim Ryan, Technology Chief at Citigroup

Goldman Sachs is focusing on measuring team velocity with AI tools rather than individual usage patterns. The bank's Chief Information Officer Marco Argenti told Business Insider that he's more interested in how quickly teams can accomplish work with AI assistance than in tracking who uses the tools most frequently.

What Does This Mean for the Future of Banking?

The scale and speed of AI deployment across Wall Street suggests that generative AI will fundamentally reshape banking operations over the next few years. JPMorgan's decision to discontinue its use of external proxy advisors and replace them with an in-house AI platform called Proxy IQ signals that banks are moving beyond using AI as a productivity tool and toward using it for strategic decision-making.

The technology is also changing what it means to work in banking. Junior bankers now need to demonstrate AI proficiency to stand out, while software engineers face new expectations around AI adoption. Goldman Sachs CEO David Solomon's message to interns encouraging them to "experiment with AI tools to enhance your teams' workflows" suggests that AI fluency is becoming a baseline expectation for new talent entering the industry.

Yet the fact that banks are still asking hard questions about returns and safety suggests that the AI revolution in finance is still in its early stages. As these institutions continue to scale their AI deployments, the focus will likely shift from simply adopting the technology to proving that it delivers measurable business value while managing risks effectively.