AI Set to ‘Hollow Out’ Banking Jobs: UK Finance Must Adapt, Warns Tech Investor
The AI Tsunami: Reshaping Banking’s Workforce
Automation’s Footprint: Numbers and Early Signs
Tech investor Mark Miller of Evantic Capital warns that large swathes of banking roles face rapid automation. Analysts put task-level automation potential in finance between 30 and 50 percent over the next decade, with some job functions exposed sooner. The earliest signs are already visible: banks deploying large language models for customer support, automated credit decisioning in lending pipelines, and accelerated trading algorithms that reduce headcount on dealing desks.
Roles most vulnerable include routine back-office processing, contact-centre agents, basic underwriting and account reconciliation. Mid-level specialist roles that rely on repeatable analysis face pressure as well. By contrast, relationship management and bespoke corporate advisory remain harder to replicate, though they will change in character.
Beyond Jobs: Implications for Financial Stability
Automation may compress banks’ revenue pools by commoditising tasks that once supported high margins. That shift can weaken profits, trim capital buffers and increase sensitivity to shocks. Other risks are model concentration and correlated exposures. If multiple firms use the same models, a single failure or adversarial attack could cascade across the sector. Operational risk from incorrect model outputs and faster decision cycles could amplify stress during market dislocation.
Regulators will need clearer guidance on model governance, stress testing for AI-driven operations and oversight of third-party model suppliers to limit systemic vulnerabilities.
The Imperative to Adapt: Embracing the AI Future
Bank leaders and policymakers must act now. Practical steps include targeted reskilling programs, redesigning operating models to shift human roles toward complex judgment and client relationships, and strengthening model risk frameworks. Investors such as Mark Miller argue that delay will force abrupt, painful adjustments. Managed well, AI can free capital for higher-value services. Mismanaged, it can hollow out institutions and magnify systemic risk. The choice for UK finance is not whether AI will change the industry. The choice is how fast institutions prepare and how effectively regulators raise the guardrails.
