Artificial intelligence (AI) agents are now moving money, and will increasingly manage financial transactions autonomously for people, threatening bank deposit profits. At the same time, large language models (LLMs) will become intermediaries to banks and emerge as new distribution channels. Finally, consumers will keep adopting AI for financial tasks, yet banks retain a trust advantage they should capitalize on. These trends appear in a new report from McKinsey and Company.
Released on November 10, 2026, the report analyzes the economic health, performance, profitability, and strategic shifts of the worldwide banking industry. It also shares new industry-wide trends and developments, drawing on a June 2026 survey of 46,000 retail banking customers across major markets.
Agents move money
Money is beginning to move based on advice from software agents, and the payment networks appear supportive of this trend: Visa’s Intelligent Commerce and Mastercard’s Agent Pay let merchants accept purchases from verified agents, PayPal agreed to put its wallet into ChatGPT’s checkout, and Stripe and OpenAI co-developed the protocol behind it.
Many analysts expect agents to start sweeping idle cash into higher-yielding accounts, a shift which could pose a threat to traditional banks. According to McKinsey, if agents directly move or indirectly guide through real-time advice even 5 to 10% of checking balances toward higher yields, industry deposit profits could drop by 20%.
But the gains from AI disruption may exceed any drop in deposit earnings. A bank that prepares for this change could already provide clients with more appealing investment products, financial guidance, and cash management tools that restore profitability, McKinsey claims.
A new distribution channel
LLMs are rising as an entirely new distribution channel in record time. While it took mobile nearly a decade to reach half of bank users, McKinsey research shows that more than 50% of users with US$100,000 in wealth are already using LLMs for financial questions, underscoring the broad and rapid disruption driven by AI.
Yet LLMs extend beyond a channel, and are increasingly acting as an intermediary between banks and clients, prompting two early strategies for banks: serving as a data source, or turning popular LLM platforms into storefronts.
In the US, OpenAI launched a set of personal finance tools for ChatGPT Pro subscribers in May 2026, enabling account links and queries like spending analysis and future financial planning, TechCrunch reported. This was done through a partnership with Plaid that allows for a connection to over 12,000 financial institutions, including Schwab, Fidelity, Chase, Robinhood, and American Express, effectively turning these institutions into data sources.
Meanwhile, institutions like Banco Bilbao Vizcaya Argentaria (BBVA) have built an app inside ChatGPT for clients to explore its products without leaving the conversation, establishing the platform as a storefront. The app was officially launched in Germany and Italy in February 2026.
Another sign of changing distribution models is the rise of generative engine optimization (GEO) as a key focus for marketing teams. GEO aims to refine brand identity, statistics, definitions, and authority so LLMs and generative search systems cite the brand in extended responses.
Together with answer engine optimization (AEO), these represent modern digital marketing and search strategies that adapt traditional search engine optimization (SEO) for AI-powered search tools and answer engines.
Consumers embrace genAI for financial tasks
McKinsey’s consumer survey shows that retail banking customers are embracing generative AI at a fast pace. Awareness of these tools is almost universal, and 23% of customers are already using them for their finance needs.

Although financial institutions are launching their own AI chatbots, popular AI platforms dominate the space. Among consumers using AI for banking, 74% rely on ChatGPT, Claude, or Gemini, about twice the share who use their bank’s chatbot today. This highlights how transactions still flow to banks, though the dialogue shaping them increasingly occurs elsewhere.
Overall, consumers are satisfied with AI, especially for simple inquiries. 59% of consumers prefer AI assistance for basic financial questions. Yet, when it comes to investment advice, 34% favor human assistance.

The study also reveals greater willingness to delegate among retail banking customers. Of AI users, 72% would allow a banking agent to perform at least one financial task. Most expect it to ask before acting, and many would limit autonomous approvals to US$500, showing caution and a need for controls.
It also found that trust remains central, with banks leading the way. 62% of AI users would rather get AI services and advice from their primary bank than from anyone else. AI users cited privacy as their greatest concern with the technology.
However, this trust advantage is conditional. About half of AI users said they might turn to a third-party financial app if their bank does not offer capable AI, implying that customers will not wait indefinitely.
The advent of institutional intelligence
Most banks have so far used AI to help employees work faster. According to McKinsey, the next phase, called institutional intelligence, will make the bank itself more capable.
Institutional intelligence will connect the data, policies, past decisions, and expert judgment that are currently scattered, gathering them so that every agent and employee can work from what the whole bank knows, a knowledge that expands on itself.
Individual productivity benefits of AI are already apparent. McKinsey’s 2026 State of AI survey showed that 80% of more than 1,700 people polled in May and June 2026 reported productivity improvements from the technology. In a benchmark by QuantumBlack, AI by McKinsey, of seven agentic systems running in production at financial and professional services firms, agents cut the time that tasks take by roughly 60%.
Building the foundation for institutional intelligence
However, these individual gains have not yet translated into broad financial impact for organizations. Just 37% of the respondents in McKinsey’s 2026 State of AI survey noted that AI improved operating profit, a figure nearly unchanged from 2025 even as more firms scaled AI technologies.
Leading banks have realized the need to redesign the work, restructure roles, and build the infrastructure to generate real value from AI. Many are now establishing the layers needed for institutional intelligence, including reliable data products like a customer 360 and a credit 360 that agents can use at scale. In McKinsey’s survey, 48% of banking executives cited data readiness as their primary AI barrier, while technology ranked as the least concern.
The second element, context, aims to map how a customer’s accounts, obligations, and regulatory status connect. Permissions are applied before information reaches a model, ensuring every agent draws from one version of the customer rather than a separate view.
Finally, the judgment element captures what the senior credit officer understands but never placed in the policy manual, captured in a format that is reusable by agents.
Featured image: Edited by Fintech News Switzerland, based on image by magnific via Magnific

