Financial Services is often seen as the poster child for AI driven efficiencies. AI excels in data-rich environments, and our industry is awash with glorious data: often unstructured, valuable, and deeply personal. AI it is postulated, is the only way to derive insights and monetize the quagmire of data we hold.

As an industry we are (rightly) tightly regulated, awash with laborious admin processes, fettered by substantial complexity, an ever-moving set of compliance obligations, and historically we can be reluctant to embrace new technology. AI is therefore a much-needed fix to the inefficiencies of our industry, which seems ripe for disruption - a theme I’ve heard repeatedly across my 25 years financial services!

On the face of it AI has indeed made significant gains in financial services. In my day job I am lucky to see innovation across 2000+ regulated firms, and over my decade of board advisory work I have never seen such focus on transformation. In some firms AI is synonymous with transformation and EY’s global study shows that 85% of financial services firms use AI primarily to reduce costs.

I see successful deployment of AI across a plethora of use cases, particularly in rule-based routine and repetitive tasks that are easy to train an LLM and automate – for example data entry, customer service, AML/KYC. AI is widely accepted as a productivity tool, and co-pilot is much loved admin assistant, for taking meeting notes, summarising documents, tracking action points, and basic research tasks. It makes sense that AI can replace a 10-person team with a bunch of automation and one human being supervising that AI. Frontier firms are moving towards AI operated workflows with a human overseeing but not performing daily tasks.

But it doesn’t just affect the lower-paid jobs in our industry, and there are a number of tasks AI can do better than a human being, as AI’s ability to learn from vast data sets outpaces human learning. Good examples of this are portfolio optimisation or financial advice for retail investors provided by “robo-advisors”.

We see early evidence of AI displacing knowledge-worker roles as AI is able to outperform humans on analytical and cognitive tasks – this has a huge knock-on effect for the traditionally well-paid middle-class roles in financial services, law, and management consultancy. This will only escalate as agentic AI becomes more sophisticated, currently generative AI is commonplace, but we are only at the foothill of what an AI agent can toddle off and do on your behalf.

The job apocalypse that never came

In reality there is huge variety in the uptake of and enthusiasm for AI across our industry, not of all of it tumultuous. Insurance and retail financial services have embraced AI for admin, but that is very different from hedge funds and trading environments who are using AI to optimise portfolios. Relationship focused business models, such as private banking and investment banks won’t be able to replace the majority of their workforce with chatbots.

In terms of seniority there is nuance to how “at risk” your job might be, whilst it might be fun to get a CEO avatar created that can write documents or partake in low level decision making it’s unlikely that AI will replace your c-suite. Even at lower levels AI doesn’t replace human beings, and ultimately AI will always need a real person “in the loop” overseeing its work.

It is more likely that the seemingly unstoppable march of AI fundamentally changes the skills we need for the future of work. Indeed, The World Economic Forum’s Future of Jobs Report 2025 projects that whilst 85 million jobs will be displaced globally by AI automation, 97 million new roles will simultaneously emerge. We haven’t got a job shortage we’ve got a skills gap, and the mid-career folk seem most at risk, with entirely new digital and AI skills required, and a crucially a significant proportion of financial services roles sit in this tier. For those reluctant to embrace AI, perhaps due to age or function, we may need to spoon feed their knowledge and skills.

Keep Calm and Carry on

On the advent of computing, it was suggested that menial work would be a thing of the past, and we’d have almost unlimited leisure time – with some of the panic around AI you’d think real humans were now superfluous at work. The current state of affairs as I see it in our industry is that AI drives productivity and the firms that don’t embrace it are not competitive – that’s not a jobs problem, rather a great example of free market economics in action.

But there are some words of caution around workforce planning in the era of AI. Care must be taken not to over-automate roles due to AI hype; a successful proof of concept not just an idea is required to show that a whole team can be successfully replaced.

For some businesses using AI is somewhat taboo – firms should be asking themselves to search questions around AI, for example: how do your customers feel about chatbots over real people dealing with their queries? How does AI fit into your ESG principles? How confident are you can deploy AI ethically? The answer to these questions should inform your HR planning, and I am concerned that I often don’t see enough strategic HR input in firm-wide AI plans. AI skills need to be measured, managed and nurtured across an organisation. Great businesses look at how to redeploy teams that are being replaced – tenure and pay is high in financial services and rehoming great people in other teams and functions makes sense.

AI: friend not foe

AI should be an enabler, removing boring work and freeing up time for more interesting strategic work. It can drive significant efficiencies, and firms investing in AI often grow faster. We are in the very early stage of AI adoption; it will be interesting to see a decade hence the changes that AI has brought to our working lives.

Katharine Wooller is chief strategist – financial services at Softcat plc