A Bloomberg Intelligence survey of European banks found research analysts, compliance and surveillance analysts, risk-modelling specialists and internal auditors among the least exposed jobs as AI spreads through the sector. The same banking machine that trims human labour at the edges is still insisting on “human oversight” while credit underwriting increasingly uses AI. The paperwork changes. The hierarchy doesn’t.
The Bank’s New Divide
Tomasz Noetzel, a senior banking analyst at Bloomberg Intelligence, said the jobs most affected by AI in banking are likely to be “in call centre, customer service staff, middle-office operations teams, retail branch employees and IT support functions”. Those are the people who keep the machine moving, answering complaints, processing routine requests, and absorbing the friction that banks prefer not to see. Noetzel said these jobs involve large volumes of repetitive work that can increasingly be handled by AI-powered assistants, but added: “That does not mean these jobs disappear overnight.”
He also said: “Demand should rise for data scientists, AI engineers, software developers … with banks expecting growth in technology and data-related roles. Clients want up-to-date information on investment portfolios which can be done with AI.” The language is neat, efficient, and very familiar. The bank keeps the high-value functions, automates the drudge, and calls it progress. Noetzel added: “Few banking jobs will be untouched, but high-judgment, specialist roles appear relatively resilient.”
The survey’s least exposed categories tell the same story from the other side of the glass. Research analysts, compliance and surveillance analysts, risk-modelling specialists and internal auditors remain relatively insulated, not because the system has become kinder, but because the people who interpret, police and stabilise it are still needed. The machine still wants its referees.
Who Gets Replaced First
The Guardian’s article also covered medicine, education and early years, law, hospitality and trades, and the pattern is the same across each sector: routine labour gets squeezed first, while the people making decisions, setting standards or managing risk are told they’re safe for now. In medicine, Hira Malik, a superintendent pharmacist and co-founder of Oushk Pharmacy, said the healthcare jobs most vulnerable to disruption by AI include medical secretaries, pharmacy support staff, prescription processing and call handling teams. She said the impact would fall on admin-led healthcare roles working with set forms, records or patient queries rather than making clinical decisions.
Malik said pharmacists, doctors, nurses and other prescribing clinicians remain far less susceptible to replacement because they carry responsibility for patient safety and treatment decisions. “AI can help organise information and flag risks, but it cannot decide whether treatment is safe or appropriate,” she said. That line lands hard. The software can sort the queue, but when the decision carries liability, the human stays in place.
Consultant plastic and reconstructive surgeon Dr Riaz Agha said: “Plastic surgery is too bespoke and too individualised. Every patient is different.” He said radiology is a speciality “which is particularly vulnerable” and added: “There have now been many studies showing that AI can interpret scans with extremely high levels of accuracy and reliability. That does not necessarily mean radiologists disappear, but their role may evolve significantly.” He said future doctors should learn how to use AI “properly and understand both its strengths and limitations”.
The Human Jobs Capital Still Needs
In education, Sharath Jeevan, founder of Oxford University’s Generational Success Lab, said: “In terms of career choices, teaching is an excellent one. Students will always need trusted adult relationships to help them learn.” Brett Wigdortz, founder and chief executive of the childcare agency Tiney, said childminding is unlikely to be taken over by technology because “people want a human being to take care of their children.” He said demand for childcare is strong, with places filling quickly, and that childminding can offer flexible, home-based work with good earning potential.
In law, Pierre Proner, the chief executive at Lawhive, said paralegal and junior lawyer roles are likely to be the most affected by AI because they often involve routine work such as document reviews, drafting first versions of legal documents, gathering information and completing forms. He said: “These are all tasks AI is especially strong at.” He added: “The roles will remain but they will just change.” Proner said AI still needs oversight. Brett Dixon, the vice-president of the Law Society of England and Wales, said automating routine tasks could create “more time and opportunities for junior lawyers to think more deeply about complex legal issues”.
Proner also said some legal areas that are less routine, such as family law or litigation, are less directly affected by AI, but that AI agents are still extremely competent at helping a lawyer prepare for a court case and run the back office of a law firm more efficiently. He said graduates should develop AI skills now and that firms are increasingly assessing candidates on their ability to use the technology, asking: “How are you using AI? Are you creating vibe-coded apps [with AI prompts]? Are you working with AI agents?” He said more people need access to the law than firms are capable of working with, and that as AI lowers the cost of delivering legal services, more jobs could be the outcome.
In hospitality, Prof Graham Miller, the academic director of the Westmont Institute of Tourism and Hospitality at Nova School of Business and Economics, said AI could reshape the distribution of jobs within hotels, shifting employment from back-office functions to front-office, customer-facing roles. He said there will always be a role for human staff in hospitality. Miller said: “I was recently in a hotel in Barcelona, and the staff there were amazing – genuinely warm, human, and welcoming. They would sit down and make you a cup of coffee. There is no way AI is doing that kind of job. That sort of human-to-human connection, which the best hotels have always delivered, should not be replaced by AI.” He said AI should make things better by handling routine tasks such as answering emails.
He said creative roles in hospitality, particularly chefs, are less vulnerable to AI than routine operational jobs, and that more routine culinary tasks such as “flipping burgers or making pizza” could eventually be automated, whereas AI is “not there yet” when it comes to producing truly innovative and creative cuisine. In trades, Brian Berry, the chief executive of the Federation of Master Builders, said AI is beginning to reshape parts of the construction sector but its impact will be uneven. He said: “Hands-on trades such as bricklaying, carpentry and plastering are less exposed to AI and continue to offer strong, long-term career opportunities,” especially for people working for small local firms.
Berry said large-scale projects could be affected in the future as some hands-on trades are automated, but implementation is still a way off. He said white-collar and administrative office job roles are being hit, including some jobs in planning and estimating. Berry said fewer than half of parents, 47%, would recommend their child take up a career in construction. He said: “That has to change. With growing demand for skilled trades and the resilience of these roles in the face of AI, construction offers a rewarding, future-proof career path, which we want more people to consider.” The future-proof pitch is doing a lot of work there. The system keeps the hands, trims the desks, and sells the rearrangement as opportunity.