The bank is reportedly considering steep reductions across parts of its U.K. wealth business. The shift shows AI moving beyond back-office automation and into work built around judgment, relationships and advice.

HSBC is reportedly considering eliminating about half of the management and specialist positions in parts of its U.K. wealth business and as much as 70% of financial-adviser roles, according to the Financial Times, as the bank moves more of its wealth operation toward digitally enabled and artificial-intelligence-supported services. Employees are already in consultation, with departures expected as early as the end of October. The scale makes this more than another bank efficiency program. Financial advice has traditionally been treated as relationship work—the kind of white-collar profession expected to remain relatively insulated as AI absorbed routine administrative tasks. HSBC is testing how much of that assumption still holds.

Chief Executive Georges Elhedery has been restructuring HSBC since taking the top job in 2024, simplifying management and pushing technology deeper into how the bank operates. HSBC told Reuters that it remains committed to delivering a high-quality wealth service through digital channels. But the proposed staffing changes show what digital transformation can mean once the technology moves from helping employees complete work to changing how many employees the institution believes it needs. Advisers are not warehouse workers or data-entry clerks. Their value proposition has historically been the ability to understand a client’s finances, risk tolerance and long-term goals, then translate those conditions into personalized guidance. AI is beginning to compete inside that layer.

SSC has been tracking this transition from several directions. In “AI Is Not Eliminating Tech Hiring. It Is Changing Where the Jobs Are,” we found that companies continue hiring even while technology changes which skills and occupations they value. In “AI Isn’t Just Replacing Jobs. It’s Rewriting What Companies Think Workers Are For,”we examined how firms can redesign organizations around AI without describing every eliminated position as automation. HSBC provides another version of that shift: the technology does not need to perform every task an adviser performs before management decides that a smaller adviser workforce can serve the same customer base.

That distinction becomes especially important in wealth management because the customer is part of the experiment. Banks have spent years moving routine transactions out of branches and onto apps. Checking a balance, transferring money or depositing a check no longer requires a person for most customers. Wealth advice is different. Decisions involving retirement, investment allocation, inheritance or major life changes can carry consequences that unfold over decades. AI can process portfolios, model scenarios and personalize recommendations at enormous scale. What remains uncertain is how much clients will value—or demand—the human layer when the decision becomes complicated, emotional or unusually consequential.

The economics give banks a clear incentive to find out. A digital wealth platform can theoretically serve more customers without increasing adviser headcount at the same rate. AI can also make existing advisers more productive by preparing client summaries, flagging portfolio changes and generating recommendations before a meeting begins. But once productivity rises, the benefit does not necessarily appear as more time for each customer. It can appear as a larger client book per adviser and fewer advisers overall. That is the pattern workers across other knowledge professions are beginning to confront: augmentation can become a head-count calculation once executives decide the augmented worker can absorb substantially more work.

HSBC’s restructuring does not prove that human financial advisers are disappearing. The bank still employs relationship managers, and wealthy clients with complex portfolios may continue to demand human attention. But it does show the threshold moving. SSC recently reported that Apple considered replacing roughly 5,000 support jobs with AI before putting that plan on hold. The HSBC case moves the experiment into a profession where trust and personalized judgment have traditionally been part of the product itself. The next phase of workplace AI may not be defined by whether a machine can completely replace a profession. It may be defined by how few people a company believes it needs once the machine becomes part of the team.