As more financial activity is delegated to agents, banks face a serious question about their role. Will they stay the main point of contact for the customer? Or will someone else own the interface, shape the decisions and capture the value?
Adoption alone won't be the decisive factor; it will more likely be the business model choice. Some banks will use agentic AI to deepen customer relationships and strengthen their position. Others risk becoming product manufacturers in the background, while AI assistants, fintechs and intermediaries take control of discovery, engagement and decision-making.
The economics of services set to change
Banks have known for years that customers want more help managing cashflow, savings, debt and goals. They've also known that help has only ever been affordable for the more affluent. Agentic AI promises to change that balance, making mass-personalised support viable – timely prompts, relevant recommendations, and practical actions for far more customers. For example, a customer could ask for help cutting their monthly spend or building towards a savings target and an agent could review their behaviour, weigh the options, act within agreed limits and keep them updated. This becomes delegated financial management.
Disintermediation is the real risk
This is happening in a market already under pressure from rising expectations, cost, regulation and digital-first competition. Here, disintermediation becomes a genuine commercial threat. If another player sits between the bank and the customer, they manage the journey, gather the data and influence the decision. The bank may still provide the regulated product but lose relevance in the moments that matter – the same way ChatGPT is starting to disintermediate shopping. That dynamic could emerge in banking, with the bank providing the product and someone else owning the relationship.
Decide what your bank wants to be
As agentic AI changes how financial activity is initiated, guided and fulfilled, banks need to decide what role they want to play. Banks that hedge their bets will struggle to make consistent choices on technology, partnerships and investment, while those that commit will find it easier to align strategy and execution.
But none of this works without the right foundations: clean, secure data; a clear division of labour between people, AI tools and agents; governance built into every workflow; and human judgement applied where it matters most.
Agentic AI will change more than efficiency in the Banking sector. It will change the very people who guide the customer, who influences the decision, and who owns the most valuable position in the chain.
Agentic AI and the reallocation of value in banking
