AI fluency is not the same as banking fluency
AI adoption in banking and asset management has by now moved well past the experimental stage and is firmly embedded in banking operations.
Consider the UK where some 75% of firms reported using AI as far back as 2024 and another 10% planned to do so within three years. The European Central Bank (ECB) reported this year that more than 85% of significant banks under European supervision use AI. In asset management, Mercer found that 91% of managers were already using AI in investment strategy or asset class research or were planning to do so. The World Economic Forum projected financial-services AI spend rising from $35 billion in 2023 to $97 billion by 2027.
This article is also available in podcast/video form. Watch the video below from our YouTube channel, or follow The Intuition Finance Digest on Spotify, Apple Podcasts, or Amazon Music.
Table of contents:
Banking AI deployment – deep but uneven
But while the depth of adoption is apparent, deployment of AI is uneven. Banks are using AI most heavily in operations, internal process optimization, cyber security, fraud prevention, customer support and coding assistance.
There is much less evidence that AI is being deployed to make fully autonomous decisions.
Looking again at the UK, a 2024 Bank of England report found that 55% of all AI use cases have some degree of automated decision-making with 24% of those being semi-autonomous – meaning that while AI models can make a range of decisions on their own, they are designed to involve human oversight for critical or ambiguous decisions. Only 2% of use cases have fully autonomous decision-making. The same report found that some 62% of all AI use cases are rated low materiality by the firms that use them with 16% rated high materiality.
AI in banking: domain expertise still critical
Given that background, it’s clear that competency in AI is already a key factor in bank recruitment. But hiring patterns indicate the while banks have keen requirement for broad-based AI literacy, this is not at the expense of domain expertise and strong capacity for judgement. It would seem the banks are aware that over-emphasis on AI fluency at the cost of underweighting banking, prudential, conduct, fiduciary, and market structure knowledge is an acknowledged risk, and that the human factor remains critical.
This principle is reinforced by regulators who are explicit that human judgement remains essential wherever accountability, ambiguity, duty, fairness, exceptions handling and severe downside risk are present.
The European Securities and Markets Authority (ESMA), for example, emphasises that management remains responsible for an organization’s decisions, regardless of whether they are taken by people or made by AI-based tools. In the Us the Office of the Comptroller of the Currency (OCC) observed in 2026 that banks’ use of GenAI and agentic AI remained largely limited to guarded use cases with human-in-the-loop accountability. The UK’s Financial Conduct Authority’s (FCA) current stance is to rely on existing conduct, governance and accountability rules rather than to create an immediate AI-specific rulebook, while using tools such as AI Live Testing to develop evidence on safe deployment.
The pitfalls of reliance on AI without the safeguards of informed human judgment are clear, illustrated already with numerous examples of embarrassing and reputationally damaging incidents – to say nothing of financial loss.
AI risks diminishing traditional banking competencies
Educators have expressed deep concern on the effects of AI on the educational development of young people. By extension, as banks automate more analytical and operational work, they risk hollowing out the very domain expertise that used to sit inside those tasks. At the same time, many employees entering banking roles may have strong communication, leadership, sales, and client skills, but limited financial services domain knowledge.
That creates a dangerous gap: people with increasing decision-making responsibility may be relying on AI-generated domain insight without having the knowledge needed to question it.
AI has an unprecedented ability to process information, summarize data and, under certain conditions, make recommendations. But in banking, the human still owns the consequences of the model’s actions, which requires judgment. That judgment depends in turn on domain fluency: the traditional banking competencies such as understanding balance sheets, credit risk, portfolio concentration, client exposure, regulation, and the knock-on effects of decisions across the bank.
Yet, while the risks of knowledge-light, AI-heavy decision-making are clear, the dramatic synergistic benefits for those who can complement banking domain knowledge with AI literacy are immense.
![[Feature Image] Compliance From obligation to business value](https://www.intuition.com/wp-content/uploads/2026/06/Feature-Image-Compliance-From-obligation-to-business-value-700x441.webp)
Frequently asked questions
Why is AI fluency not the same as banking fluency?
AI fluency means understanding how to use AI tools, prompts, automation, and model outputs effectively. Banking fluency is different because it depends on financial services knowledge, judgment, regulation, risk, client exposure, and accountability. The article argues that banks need both. AI can process information and make recommendations, but people still need domain expertise to question outputs and understand the consequences.
How widely is AI being adopted in banking and asset management?
AI adoption is already well established across banking and asset management. The article cites UK firms using AI, European banks under supervision deploying AI, and asset managers using or planning to use AI in investment strategy or research. It also notes that financial services AI spending is projected to rise significantly, showing that AI is no longer experimental in the sector.
Where are banks using AI most heavily today?
Banks are using AI most heavily in operations, internal process optimization, cyber security, fraud prevention, customer support, and coding assistance. The article makes clear that adoption is deep but uneven. AI is widely embedded in practical and guarded use cases, but there is much less evidence that banks are using it for fully autonomous decision-making across high-risk banking functions.
Why does human judgment remain critical in banking AI?
Human judgment remains critical because accountability, ambiguity, duty, fairness, exceptions handling, and severe downside risk still sit with people and institutions. The article notes that regulators expect management to remain responsible for decisions, even when AI tools are involved. In banking, AI can support decisions, but humans still own the consequences of those decisions.
What risk does AI create for traditional banking competencies?
As banks automate more analytical and operational tasks, they risk weakening the domain expertise that employees used to build through those tasks. The article warns that people may gain decision-making responsibility while relying on AI-generated insight they cannot properly challenge. This creates a dangerous gap between AI-assisted work and the banking knowledge needed to judge whether that work is sound.
What banking knowledge is still needed alongside AI literacy?
The article highlights traditional banking competencies such as understanding balance sheets, credit risk, portfolio concentration, client exposure, regulation, and the knock-on effects of decisions across the bank. These areas help employees question AI outputs and apply judgment. The strongest value comes when banking domain knowledge and AI literacy work together rather than one replacing the other.
Build the banking fluency your teams need to use AI with confidence, sound judgment, and regulatory awareness across modern financial markets with Intuition Know-How.
Click here to learn more, or fill out the form below.
