AI News
20 Aug 2026
Read 10 min
How Wall Street Banks AI Strategy Is Saving Billions
Wall Street banks AI strategy cuts costs and boosts firmwide productivity across trading risk and ops.
Wall Street banks AI strategy is shifting from hype to hard savings. JPMorgan, Goldman, Citi, Wells Fargo, Bank of America, and Morgan Stanley are spending billions and rolling out assistants, agents, and coding copilots. Early wins include faster software delivery, leaner operations, and new client tools—while leaders still watch costs, risk, and returns.
Big banks want speed, safety, and clear paybacks from AI. They are putting the tools in every team, not just in labs. They track productivity, move workflows to AI agents, and reset goals for engineers. Results now show up in coding hours saved, fewer manual steps, and smarter client service.
Wall Street banks AI strategy: What leaders are doing now
JPMorgan Chase
JPMorgan backs AI with nearly $20 billion a year for tech. Jamie Dimon says a $2 billion AI push has already paid for itself through savings. The bank tracks developer use of GitHub Copilot and expects engineers to adopt AI to “drive excellence.” It counts about 1,000 live use cases, from fraud checks to marketing to meeting notes. A firmwide genAI platform now reaches 200,000+ staff. Leadership also reshaped data and analytics to move from building infrastructure to hitting business goals. In asset management, a new platform called Proxy IQ will guide US proxy voting in-house, replacing outside advisors.Goldman Sachs
Goldman is spending about $6 billion on tech this year. CEO David Solomon wants more, but he must protect returns. An internal program, OneGS, aims to use AI to cut friction, slow hiring, and trim some roles. Goldman launched an AI assistant for employees and works with Anthropic on agents for trade and transaction accounting and for client onboarding. CIO Marco Argenti measures team velocity with AI, not just how often a person clicks a tool.Citigroup
Citi chases buy-in first. The bank trained 4,000 “AI stewards” and says nearly 90% of employees now use AI. It rolled out an AI wealth advisor and is scaling agents. Tech chief Tim Ryan favors “metrics and pride,” and he avoids close tracking of every prompt. Teams are urged to pick the lowest-cost model that gets the job done. Citi also refreshed leadership to speed its AI push.Wells Fargo
Wells uses a hub-and-spoke model: a small central AI team connects with genAI leads in each business. AI agents power some of the highest-impact tasks. The bank did not mandate AI use when it spoke in March, but momentum is rising. It launched an “AI teammate” for financial advisors to lift daily activity. CEO Charles Scharf says engineers are up to 35% more productive with genAI. Wells also hired AWS veteran Faraz Shafiq to lead AI products and solutions.Bank of America
Bank of America devoted $13 billion to tech in 2025 and says AI runs across consumer and institutional lines. It has approved 300+ AI/ML cases and fully deployed at least 34 genAI uses. Wealth, investment banking, and engineering teams use specialized tools. Clients have engaged with its virtual assistant, Erica, more than 3.2 billion times since 2018. The bank uses models to issue daily markets notes and to help draft pitch books.Morgan Stanley
Morgan Stanley partnered early with OpenAI and equipped advisors with genAI starting in 2024. A developer tool, DevGen.AI, saved 280,000 hours in six months by parsing legacy code. The firm also highlighted Parable for data summaries and LeadIQ for matching clients to advisors. A 2025 intern survey showed 72% used ChatGPT daily or several times a week, signaling how quickly AI is becoming standard.What the spending buys—and what it saves
Banks now steer money toward tools that raise speed, cut manual work, and reduce risk. The pattern is clear across leaders.- Engineering velocity: Copilots help write and fix code. Morgan Stanley’s DevGen.AI saved 280,000 hours. Wells Fargo cites up to 35% productivity gains.
- Process automation: Goldman and Anthropic built agents for accounting and onboarding. JPMorgan applies AI to note-taking and marketing to remove low-value work.
- Risk and controls: JPMorgan runs hundreds of models, including for fraud and security. Broad deployment brings faster alerts and better monitoring.
- Client experience: Bank of America’s Erica handled 3.2 billion interactions. Citi and Morgan Stanley ship new advisor tools, while Wells adds an AI teammate.
- Operating model: Goldman’s OneGS and JPMorgan’s reorganizations align teams, budgets, and governance around AI outcomes.
Playbook takeaways any bank can copy
- Start with enterprise platforms and guardrails. Give every team access, but keep data safe and models governed.
- Measure outcomes, not clicks. Follow Goldman’s lead by tracking team velocity and release rates, not only individual usage.
- Build bottom-up momentum. Citi’s AI stewards and training unlock use cases in every line of business.
- Adopt AI agents for workflows, not just chat. Use agents to reconcile trades, onboard clients, draft documents, and route leads.
- Right-size models to the task. Encourage the lowest-cost model that meets the need, as Citi does.
- Reskill at scale. Set expectations (like JPMorgan) that engineers use AI and prove impact against clear goals.
- Partner where it speeds impact. Work with model labs (e.g., Anthropic, OpenAI) and cloud leaders to accelerate delivery.
- Publish the wins. Track hours saved, cycle-time cuts, error-rate drops, and client adoption—then reinvest savings.
(Source: https://www.businessinsider.com/wall-street-banks-ai-strategy-spending-jpmorgan-citi-goldman-2026)
For more news: Click Here
FAQ
Q: What are the main goals driving Wall Street banks’ AI investments?
A: The Wall Street banks AI strategy focuses on increasing speed, improving safety, and securing clear paybacks by embedding AI tools across teams rather than keeping them confined to labs. Banks aim to reengineer workflows, raise engineering velocity, and reduce manual work to realize cost savings and faster delivery cycles.
Q: How much are top banks spending on AI and technology?
A: Examples in the article show large budgets: JPMorgan has a nearly $20 billion annual technology budget and said a $2 billion AI push has already paid for itself, Goldman put about $6 billion behind tech this year, Citi oversees roughly $12 billion, and Bank of America devoted $13 billion to tech in 2025. These figures illustrate the scale of investment backing many firms’ AI initiatives.
Q: What early wins have banks reported from their AI deployments?
A: Early wins from the Wall Street banks AI strategy include faster software delivery, leaner operations, and new client tools, such as Morgan Stanley’s DevGen.AI saving developers about 280,000 hours in six months and Wells Fargo reporting up to 35% productivity gains for engineers. Banks also cite fewer manual steps and high client engagement, with Bank of America’s Erica handling more than 3.2 billion interactions since 2018.
Q: How are banks measuring the impact and productivity of AI?
A: Banks are shifting to measuring outcomes rather than simple usage metrics, tracking team velocity, release rates, hours saved, and business results. For example, Goldman focuses on teams’ velocity, Citi avoids granular prompt tracking while JPMorgan monitors developer Copilot usage and sets new objectives for engineers.
Q: How are banks organizing teams and governance for AI deployment?
A: Firms use a mix of centralized governance and business-line embedding, evident in Wells Fargo’s hub-and-spoke model and JPMorgan’s restructuring of its chief data and analytics office to move from infrastructure to business initiatives. Banks have also reshaped leadership and operating models, with Citi appointing new AI leaders and Goldman running cross-bank initiatives like OneGS.
Q: What risks and concerns do banks and analysts highlight about AI?
A: Analysts and executives continue to raise questions about returns, safety, and controls as firms scale AI, emphasizing the need to justify large investments and manage operational risk. In response, banks run models for fraud and security and emphasize governance, guardrails, and measured rollouts to balance scale and risk.
Q: What specific AI tools or platforms have individual banks rolled out?
A: Banks have deployed a range of tools, from JPMorgan’s proprietary genAI platform and Proxy IQ to Bank of America’s Erica virtual assistant and Morgan Stanley’s DevGen.AI, Parable, and LeadIQ. Goldman has an internal AI assistant and works with Anthropic on agents, while Citi and Wells have been scaling advisor tools, agents, and an “AI teammate” for financial advisors.
Q: What practical playbook steps does the article recommend other banks copy?
A: The article recommends starting with enterprise platforms and guardrails, measuring outcomes not clicks, building bottom-up momentum through training and stewards, adopting AI agents for workflows, right-sizing models to tasks, reskilling engineers, partnering with model labs and cloud providers, and publishing wins. These steps summarize the Wall Street banks AI strategy for moving from experimentation to measurable savings and faster delivery cycles.
Contents
