The Financial Times reported on Wednesday 7 October that HSBC is proposing sweeping reductions inside its UK wealth-management operation as it pushes more of the service towards digital tools and artificial intelligence. Reuters, which carried the report, said the plan could remove about half of management and specialist roles and as many as 70% of financial-adviser posts in the affected business.
Those percentages are striking. They are also reported proposals, not a completed redundancy announcement with a published headcount. HSBC does not disclose how many people work in the particular UK wealth unit, and the FT cited people familiar with the plan. The bank is in consultation, according to the report. That distinction matters to employees, customers and anyone tempted to turn a complicated reorganisation into a tidy headline declaring that a robot has already collected everybody’s pass.
HSBC told Reuters that it remained a long-established UK wealth manager and was continuing to evolve towards more digitally enabled products and customer journeys. It did not confirm the reported percentages in that statement. The FT said affected staff could leave by the end of October.
What is actually being changed?
Wealth management sits between ordinary retail banking and the bespoke service offered to the very richest clients. Advisers help customers decide how to invest, plan retirement, use tax allowances and balance risk. Managers, researchers, compliance specialists and support teams sit behind those conversations.
Digital wealth platforms can already collect information, classify a customer’s risk appetite, suggest a model portfolio and produce routine documents. Generative AI can summarise meetings, retrieve product information, prepare correspondence and help advisers search large rule books. Those capabilities can reduce administration and allow one employee to serve more clients.
They do not remove the bank’s legal responsibilities. Financial advice must be suitable for the individual receiving it. A customer facing retirement, bereavement or a volatile market may need explanation and judgement rather than a glossy dashboard. If an automated system recommends the wrong product, the customer cannot sensibly be told that the algorithm has moved departments.
That is why the most important unanswered figure is not merely how many posts disappear. It is the adviser-to-client ratio after the reorganisation, the availability of human help and the controls applied to AI-generated work. Cutting 70% of advisers, if the reported proposal is implemented at that scale, would be a profound service redesign rather than a software upgrade.
The AI claim needs careful handling
HSBC chief executive Georges Elhedery said at an investor event in May that staff should embrace AI-led change and that generative AI would destroy certain jobs. He has made the technology central to the bank’s strategy since taking charge in 2024.
That does not prove every role in this proposal is being replaced by a single AI system. Large restructurings usually combine automation, simpler products, changed sales channels, management delayering and cost targets. “AI” can describe a genuine technological shift, but it can also become the fashionable label attached to decisions that would once have been called consolidation.
The useful test is evidence. Which tasks will machines perform? Which decisions remain with regulated humans? How will accuracy, bias, data security and customer outcomes be measured? Will savings improve prices and access, or mainly improve the cost-income ratio? Until HSBC publishes more detail, confident answers would be forecasts dressed as facts.
Britain has already seen workers paying for AI tools themselves and sometimes using them without their employer’s knowledge, as our report on UK workers spending their own money on AI explained. This proposal moves the issue from informal workplace experimentation into the formal design of a regulated financial service.
What it means for customers
Digital service is not automatically inferior. Routine investing can become cheaper and faster when software removes repetitive paperwork. Customers who prefer an app may welcome progress tracking, quicker responses and clearer information. Human advisers also make mistakes, and a well-controlled system can identify inconsistencies that a rushed person misses.
But wealth advice is not a weather widget. People disclose salaries, debts, dependants, health expectations and long-term plans. The bank must protect that data and make clear when a customer is interacting with automated content. It must also provide a route to a competent person when circumstances do not fit the standard journey.
Vulnerable customers are an obvious concern. Digital exclusion can be financial, physical or cognitive. A service designed around confident smartphone users can quietly become harder to use for people who most need help. Regulators should therefore look at outcomes by customer group, not accept an average satisfaction score as proof that nobody has been left behind.
The Financial Conduct Authority’s Consumer Duty requires firms to deliver good outcomes for retail customers. It does not provide an exemption because a chatbot produced the document faster. HSBC will remain responsible for communications, support and suitability across whichever mixture of employees and systems it chooses.
What it means for work
The reported plan is unusually important because financial advice is skilled, regulated and relationship-based. The comforting version of the AI story says automation removes drudgery while people move into richer work. If large adviser reductions happen, that version needs updating: the technology may increase the number of customers one adviser can cover, while reducing the total number employed.
Consultation should test alternatives rather than merely explain a settled outcome. Employees and representatives will want the role counts, selection criteria, retraining offers and geographic impact. The bank should publish enough information for the public to distinguish genuine productivity investment from a blunt reduction wrapped in futuristic language.
There is also a pipeline question. Senior advisers do not appear fully formed. If entry and mid-level work is automated, banks still need a way to train the humans who will handle complex cases, supervise systems and take responsibility when the model is uncertain.
The OutOut verdict
Used properly, AI can give an adviser better information and give a customer faster service. Used as a boardroom adjective, it can make an old-fashioned headcount reduction sound like a tour of tomorrow.
HSBC is entitled to redesign its business, and no bank should preserve pointless processes as a museum exhibit. But the proposed scale reported here demands more than promises about digital journeys. Employees deserve clarity; customers deserve human escape routes; regulators deserve measurable safeguards. Artificial intelligence may write the meeting summary. Accountability still needs a name badge.
