With all the media noise around AI-related job reductions in banking, it’s no wonder that many employees are fearful of their future and have concerns about AI taking their jobs. But the more I speak to bankers, technologists and transformation leaders in interviews and on panels at events, the more I come around to thinking that the story of AI in banking is not that machines are becoming more capable, but that people are becoming more essential.

At first glance, that sounds backwards. We are living through a period in which generative AI and agentic AI are being rolled out to automate analysis, draft content, surface recommendations, summarise interactions and speed up decisions. Yet the practical effect of all this, at least from where I sit, seems to be a renewed appreciation for the things only people can really do well: exercise judgement, build trust, communicate clearly, recognise nuance, and respond with empathy when the stakes are high.

That impression is not just anecdotal; it is increasingly supported by research by McKinsey, for example.

AI is changing tasks faster than it is replacing human value

One reason this idea feels more convincing now is that the evidence is starting to separate tasks from roles. The World Economic Forum’s Future of Jobs Report 2025 shows that AI and information processing are among the most transformative forces reshaping work by 2030, but it also says the skills rising alongside this shift include creative thinking, resilience, flexibility and agility. In other words, as technology becomes more powerful, the premium on human adaptability rises with it.

McKinsey makes the same point even more directly. Its view in the article linked above is that AI will transform tasks, but human work will endure, and that the organisations capturing the most value will be the ones that invest in people as a core asset rather than simply deploying tools. I think that distinction matters massively for banking. A bank can automate a workflow, but it still needs someone to interpret exceptions, challenge recommendations, explain outcomes to clients, and take responsibility for the consequences.

During Sibos 2025, I was intrigued by a comment made by Leigh-Ann Russell, Global Head of Engineering and CIO at BNY, who told the audience that they already had over 100 non-human (AI) employees, with their own login credentials, email and Microsoft Teams accounts who joined meetings and carried out tasks delegated by human managers. Following up afterwards, I found further information on their website that states: “By putting AI in the hands of everyone at BNY, we intend to develop fluency and create capacity for our people to focus on higher-value work”.

That’s why I don’t see AI as a straightforward substitute for human capability. I see it more as an amplifier. It reduces the time spent gathering information, searching policies, producing first drafts and spotting patterns. That gives bankers more room to focus on what clients and colleagues remember: the quality of the conversation, the clarity of the advice and the confidence that a real person is exercising sound judgement.

The market is already signalling what matters next

I also find it telling that the market is signalling a rise in precisely the kinds of skills people often label as “soft”, even though there is nothing soft about them in practice.

The UK picture is especially interesting. Reporting on LinkedIn’s 2025 UK Skills on the Rise findings shows that relationship-building and strategic thinking are at the top of the list, with AI literacy in third place. That combination says a lot. It does not suggest that human skills are replacing AI skills, or vice versa. It suggests that the winners will be the people who combine both. LinkedIn’s UK data also points to a growing shortage of soft skills, with 45% of UK hirers saying that finding candidates with the right soft skills is one of their biggest hiring challenges.

For banking leaders, that should be a wake-up call. If everyone is investing in AI capability, then the differentiator may become the quality of the human layer around it.

In practice, AI often makes human interaction more valuable, not less

What I find most interesting are the examples from banks themselves, revealed during the interviews and panels I have been engaged in this year. They are not only using AI to cut costs. Increasingly, they are using it to make employees more effective in moments that still require a human touch (as per this great example provided by Deepak Pandey in an interview earlier this year).

We previously reported on how BNP Paribas had rolled out it’s AI assistant “Yara Invest” to private banking teams to help synthesise investment reports, translate documents, and generate tailored financial insights.

One key advantage of Yara Invest is its ability to customise content based on individual client profiles, ensuring that investment information is presented that aligns with each client’s expertise level and preferences, and to do this at a much faster pace.

That is a good example of what I mean by helping humans be more human: the technology handles the admin and pattern recognition, while the employee can focus more fully on the conversation itself.

Marco Li Mandri (Head of Analytics at ING) argues that AI cannot yet handle banking aspects that rely on judgment, accountability, and trust. He points to big, high-stakes decisions like choosing the right mortgage or selecting an investment strategy, where the “human advice” element still matters most, not just for reassurance but because these decisions are personal, regulated, and often nuanced.

The value of AI is not framed as replacing the advisor; it is framed as making the advisor better prepared, better informed and more responsive to each client’s unique needs.

Why empathy and judgment rise when information becomes abundant

There is a broader reason I think this is happening. When information is scarce, expertise is often about access. When information becomes abundant and instant, expertise shifts towards interpretation.

AI is brilliant at surfacing options, detecting anomalies and synthesising documents at speed. But in banking, the hardest moments are rarely about information retrieval alone. They are about trade-offs. Should a vulnerable customer be pushed through a standard process, or should someone step in? How should a relationship manager explain a difficult credit decision? How should a fraud analyst balance pattern recognition with fairness and proportionality? These are not purely technical questions. They are moral and relational ones.

Capgemini makes a compelling case that by automating routine enquiries, AI allows bank agents to focus on complex, high-value tasks requiring empathy, creativity and judgement, and to deliver more personalised service at scale. I think that gets to the heart of it. AI does not make empathy obsolete. It makes empathy more visible by stripping away some of the lower-value work around it.

The real competitive advantage may be human quality at scale

My own conclusion is that AI will not make banking less human. Used well, it could make the human dimension of banking more intentional and more valuable.

Reframing AI as “Augmented Intelligence” instead of “Artificial Intelligence”.

That will not happen automatically. Banks will need governance, training, accountability and better organisational design. They will need people who can question outputs, communicate clearly with customers, and apply judgment when the model is directionally useful but contextually incomplete. But if they get that balance right, AI could do something surprisingly positive: it could free people from the mechanical parts of work and push them towards the parts that clients value most.

So, to answer my own question in the title of this article, I do think artificial intelligence can help humans be more human.

Not because the technology becomes warm, wise or trustworthy on its own, but because it can give bankers more time, more context and better prompts for action. And when that happens, empathy, judgement, relationship management and communication stop looking like secondary qualities.

They become the main event.

Join Belgium's Biggest Banking Conference in Brussels on May 28, where AI and the future of banking will feature heavily on the agenda, for insights and inspiration from industry leaders from the Benelux and beyond.