On October 7 the Financial Times reported that HSBC plans to cut about half of the management and specialist roles in its UK wealth management business, and to shrink its human financial advisers by about 70%. The paper’s sources say AI is being used to cover the affected work. HSBC has not confirmed the figures. Its response, as Computer Weekly relayed it, was that the bank is “continuing to evolve to deliver more digitally enabled products and journeys.”
Everything below rests on that one report and the follow-ups that cite it. We label it reported, not announced, and we do not know which AI tools are meant to replace which roles. The report doesn’t say.
What the report says
Two percentages and a deadline. Roughly half of the unit’s management and specialist positions. About 70% of its human advisers. Staff are in consultation now, and the people whose roles go are due to leave before the end of October, which is 24 days from the day the story ran.
There is no headcount. HSBC doesn’t publish a UK wealth figure, and the FT’s only anchor is that UK relationship managers number in the hundreds. So the honest reading is that this is a cut of hundreds, perhaps more, inside one division, not a bank-wide number. A source quoted by Computer Weekly called the programme “deep, wide and brutal.” That is a description of the sourcing, not a count.
The fuse is the story
Set it beside the last time a global bank put AI job losses on paper. On May 19, Standard Chartered gave a number and a date at its investor day: 7,000-plus corporate-function roles, about 15% of a 52,271-person back office, by 2030. That is a runway of more than four years.
HSBC’s version has a runway of under four weeks. One is a target announced from a stage. The other is a consultation already running, reported before the bank chose to say anything. Mizuho’s up-to-5,000 cut, spread over a decade, sits at the slow end of the same cohort. If the FT is right, HSBC is at the fast end, and it picked the one part of the bank where the customer is a person across a table.
HSBC’s chief executive had flagged the direction. The day after Standard Chartered, on May 20, Georges Elhedery told staff at the bank’s investor day that generative AI would destroy certain jobs and create new ones, as The Standard reported. He gave no figure then. This is the first time a figure has attached to the sentence, and it came from a leak.
The AI trail leading to wealth
Computer Weekly lists what HSBC has built since March. It named its first chief AI officer in March. A June Google Cloud deal adds more than 200 AI use cases over two years, on top of roughly 600 applications already running there. In July it announced an AI centre of excellence in Singapore, aiming to hire 100 AI specialists, with an initial focus on applying AI to customer wealth conversations.
That last item matters. The unit being cut is the unit whose conversations the new centre was set up to work on. Our read, not a reported fact: the sequence reads as a build-out first and a headcount decision after, which is how the Standard Chartered case ran too. The FT also noted that Evident’s annual AI Index, published the day before the story, ranked HSBC 11th, down from 8th. A bank slipping in an AI league table has a reason to show movement.
What it means for the people in the role
The roles named are management layers and specialists in a wealth business, and the advisers who sit under them. The exposure mechanism, as reported, is not a chatbot replacing a clerk. It is AI absorbing the part of an adviser’s day spent on the repeatable parts of a client conversation, which leaves fewer advisers carrying larger books.
Two timing points. First, the consultation window is short, so anyone in the unit has weeks, not quarters, to decide whether to move inside HSBC, leave with a client relationship, or leave without one. Second, a 70% reduction in advisers does not remove the function. It concentrates it. The 30% who remain are the ones whose client relationships are the product. The layer above them is also being cut, by half, and that is the less usual part: in most bank AI plans the managers stay and the front line thins. Here both thin.
For advisers elsewhere in UK retail wealth, the signal is the ratio, not the bank. A large firm has now been reported to plan for three advisers out of ten. If that holds, other firms’ internal models will use it.
What we can’t tell from the report
Three gaps limit how far this can be pushed. There is no base: without the size of the unit, 70% could be a few hundred people or several times that. There is no split between redundancy and redeployment, so “leave before the end of October” may include moves to other HSBC roles. And there is no stated role for AI beyond the FT’s sourcing: the bank’s own words, “digitally enabled products and journeys,” are a product description, not a staffing plan.
What the report does establish is direction and speed, from a source close enough to the consultation to name both percentages. That is enough to log it. It is not enough to cite as an HSBC disclosure, and the Monthly Report should not.
What to watch
Whether HSBC confirms or corrects the numbers. Whether Unite or Accord publish a position, since Computer Weekly’s account carried no union comment. Whether any other UK bank is reported to be running a similar consultation. And the bank’s third-quarter results, which are where a restructuring charge for this would show up.
