Skip to main content
VTE

AI by sector

AI in Banking and finance

Banking was among the first sectors to adopt AI at scale, and today it's one of the most exposed: much of its work —analysing information, spotting patterns, drafting reports, handling queries— is exactly what models do well. But approving a risk, holding a client's trust and answering to the regulator remain deeply human. The question isn't whether AI enters banking —it already has— but which part of your work it reorganises.

What's already happening

  • Real-time fraud and money-laundering detection over millions of transactions.
  • Model-assisted credit scoring and risk assessment.
  • Chatbots and assistants for first-line customer service.
  • Back-office automation: KYC, reconciliations, regulatory reporting.
  • Analyst copilots: summarising reports, drafting and exploring scenarios.

Where the line is

AI already does

  • Screen and classify transactions
  • Draft report first cuts
  • Answer frequent queries
  • Process and extract data from documents

Stays human

  • Approving (and owning) a risk
  • The trust relationship with the client
  • Accountability to the regulator
  • The exception and judgment under ambiguity

Key occupations

Open each occupation to see its exposure, what changes and what to do.

What to do: the 3 A's

Automate the routine

Hand AI your reconciliations, routine reports and first-line replies. It's the repeatable work eating your time without being where you add value.

Augment your judgment

Use copilots to analyse faster, summarise dense documentation and draft first cuts —always with your judgment validating the output before it reaches a client or a risk committee.

Anticipate what's next

Move up to where AI can't reach: risk judgment, client relationships and regulated decisions. That's where a banking professional becomes more valuable, not less.

The number

Typical banking occupations score between 65 and 82 out of 100 on AI exposure (mean ~71), among the highest in the economy.

Own aggregation over the AIOE index (Felten et al., 2021) and “GPTs are GPTs” (Eloundou et al., 2024).

Frequently asked questions

Will AI replace banking employees?
Almost never all at once. Exposure measures task overlap, not replacement: usually what changes is what you do and how —toward risk, relationships and decisions— not whether your role exists. Those who use AI well tend to gain weight, not lose it.
Which banking roles change the most?
The screen-intensive, repeatable-task ones: back-office, analysis, first-line service. Those resting on relationships, risk judgment and regulated accountability change less at their core, even as their day-to-day leans more on copilots.
What about investment banking or advisory?
AI speeds up analysis, documentation and scenario generation, but the client relationship, reading the context and owning the recommendation stay human. The advisor gains time for what truly differentiates.

Your next step

See all sectors