ISSUE 15 · FINANCE & STRATEGY
🔍 THROUGH THE LENS
Last Thursday morning, a BBC headline stopped me: Barclays is cutting about 5,000 jobs this year, around 5% of its workforce, as it uses AI and other technology to lower its fixed costs. The bank said about 60% of the cuts will fall in support functions, developers included.
In one of my previous roles at a bank, I worked on service simplification: taking steps out of customer journeys and finding out why customers called or came into a branch. Every call and every branch visit costs a bank money. So the aim was to remove the steps customers didn't need. Fewer steps meant fewer calls and fewer branch visits, and that meant fewer people in both.
So the number itself didn't surprise me. Barclays cut a similar number of roles in 2023, before AI was the headline. What caught my attention is the reason now given, and where the cuts land: not branches and call centres anymore but the support functions, the teams that run a bank behind the scenes - back office.
Which raises a question for anyone in a support role. Is AI already doing this work, or is it the newest name for cost-cutting banks were doing anyway?
Unite, the union that represents bank workers, had its answer. Its general secretary, Sharon Graham, said Barclays is "using AI as an excuse to discard its loyal and long-suffering workforce". The bank's chief executive sees it differently, as you'll read below.
Here's what the numbers say, and what it means for your job.
🧭 THE SHIFT
Banks are using AI to cut costs.
What happened: Barclays isn't the first. In May, Standard Chartered said it will cut about 7,800 roles by 2030, more than 15% of its support functions, as part of a new strategy built on AI.
The details:
Costs first, revenue later. In July, a D.A. Davidson survey of US banks found they expect AI to cut expenses by about 4.5% to 5% over the next two to three years, against 2.5% to 3% revenue growth. Only 11% said the savings are measurable.
Europe too. In June, Banco Santander set a target of more than €1 billion in AI value between 2026 and 2028, with more than half expected from cost cuts.
Cuts before redesign. On 8 October, Gartner found that 63% of senior procurement leaders expect AI to significantly improve their function within three years, but only 25% have redesigned jobs because of it. Procurement is a support function too.
💼 WHAT THIS MEANS FOR YOUR CAREER
In the digital transformation years, the front line took most of the cuts: branches and call centres. This wave is reaching the back office.
Barclays' chief executive, C.S. Venkatakrishnan, put it plainly: "AI is the icing on the cake not the cake. Because you need to do the infrastructure build to get the benefits of AI."
The cake is everything underneath: the systems, the processes and the people who run them.
As anyone in a large organisation knows, the rebuild is the hard part, not the tech or AI: old systems, unclear ownership, slow sign-off. That's why there won't be one big AI moment. Change will come team by team, a few hundred jobs at a time, sometimes a few thousand. Once the rebuild is done, it will speed up.
Finance isn't exempt. A Bloomberg Opinion analysis of ONS data last week showed UK hiring for finance roles is still well below pre-pandemic levels. And in October, investors at a16z wrote that the leanest finance teams now aim for about 2% of company headcount, down from around 5%.
🎯 GET READY
Last week I asked on LinkedIn what bothers you most about AI and your career. Most of the handful of answers said the same thing: not sure which skills to learn. So here's where I'd start.
Junior: Learn where your numbers come from: the ledger, the systems, the SQL behind your reports. Let AI draft your report, then prove every line against the data. Volunteer for the breaks and exceptions. That's where judgement gets built now.
Mid: Become your team's finance engineer. Automate one month-end or forecasting step in a tool your firm approves, and build in the audit trail from day one: what the tool did, and who signed it off.
Senior: Map your team task by task, then set the rules before agents arrive: what an agent may draft, what a person must approve, and how it's evidenced. Measuring AI's value is becoming a finance job too.
📰 ALSO WORTH KNOWING
Elsewhere: Kazakhstan's AI bet: On 23 October, Nvidia's CEO Jensen Huang made his first visit to Kazakhstan, closing the country's AI Month in Astana. He said the country can "transform the energy into intelligence and export intelligence," The Astana Times reported.
Why it matters: The infrastructure that slows AI inside banks is being built at scale elsewhere. And Huang's own advice fits this issue: "you invest in people, not companies."
🏛️ Regulation: On 15 October, the Financial Stability Board published its final 12 sound practices for adopting AI responsibly in financial firms. They put boards and senior managers in charge of AI governance and include human oversight.
Why it matters: They're not binding, but supervisors will use them. If AI takes over support work, someone still has to answer for what it does.
Wondering where your role is heading?
Book a free 30-minute career chat. Pick a time that suits you.
—————