Morgan Stanley has doubled its forecast for AI-driven job losses across the European banking sector, estimating that as much as 20% of total banking employment could be eliminated by 2030 as lenders push generative-AI tools into back-office, risk and compliance workflows.
The revised figure, reported by Bloomberg on Thursday, lifts the estimate to roughly 400,000 jobs from the 200,000-job, 10% projection the bank published in January. The doubling is the part worth pausing on. Five months ago, Morgan Stanley analysts argued AI deployment across the European banking sector would translate into around 200,000 cumulative role eliminations by the end of the decade, concentrated in back-office, KYC-and-AML compliance, and middle-office risk-monitoring positions.
The May revision keeps the same functional concentration but scales the headline number up substantially. What changed in five months, on the bank’s framing, is the pace at which individual European banks have begun publicly committing to AI-led restructuring, alongside earnings-call signals that productivity gains from generative-AI deployment are materialising faster than even bullish 2025 forecasts had assumed.
The bank-by-bank evidence is concrete. ABN Amro announced in November 2025 that it would cut roughly 20% of its full-time workforce by 2028, primarily through automation. HSBC has committed to cutting around 20,000 jobs as AI absorbs back-office work, with chief executive Georges Elhedery explicitly framing the reductions as productivity-led rather than cost-driven. UBS, which is still working through the Credit Suisse integration, has begun a fresh round of cuts in Switzerland that the bank expects to deliver roughly half of its targeted $10bn cost-saving programme through 2026. Société Générale chief executive Slawomir Krupa said in March that “nothing is sacred” in the French bank’s cost-reduction programme. BNP Paribas, the eurozone’s largest bank by assets, has paired its AI-driven cost work with an unusually visible Mistral partnership on the foundation-model side.
Regulatory Hurdles and Workforce Recomposition
The regulatory question is whether European labour law actually permits bank-by-bank reductions on the scale Morgan Stanley is now projecting. France, Germany, the Netherlands and Spain all have works-council and collective-bargaining structures that make rapid workforce cuts substantially harder than US-style at-will layoffs. The 20% figure, on Morgan Stanley’s framing, assumes the cuts are achieved primarily through attrition, early retirement, and managed exit programmes over a five-year window rather than through mass redundancy. Whether the regulatory frame holds if cost pressure intensifies further is a separate question.
The ECB’s position is itself relevant. The European Central Bank’s supervisory arm has been explicitly pushing eurozone banks to accelerate their AI cyber-security posture in response to threats from Anthropic’s Mythos and similar adversary tools, which structurally requires more technology-and-data-engineering capacity inside banks even as the back-office headcount declines. The net result, on the Morgan Stanley analysis, is that the workforce shift will be a structural recomposition rather than a flat reduction: data engineers, AI-platform operators and model-risk specialists in, traditional compliance officers and back-office processors out.
Historical Context and Productivity Trajectory
The financial services industry has long been at the forefront of automation. From the introduction of ATMs in the 1970s to the rise of online banking and algorithmic trading, technology has consistently reshaped employment in the sector. However, the current wave driven by generative AI is fundamentally different. Unlike previous automation that targeted specific repetitive tasks, generative AI can understand, summarise, generate, and analyse human-like text, enabling it to take over complex decision-making processes in compliance, risk assessment, and customer service. The pace of adoption is unprecedented: banks are deploying chatbots for client queries, AI systems for anti-money laundering checks, and machine learning models for credit risk evaluation. These tools reduce the need for human intervention, leading to significant headcount reductions.
Productivity gains from generative AI are materialising faster than many analysts anticipated. A study by Accenture in early 2025 found that AI-assisted workers in banking could complete tasks up to 40% faster, with error rates dropping by half. This efficiency is driving shareholder pressure for cost-cutting, especially in a low-interest-rate environment where revenue growth is constrained. European banks, already grappling with negative rates and regulatory capital requirements, see AI as a lever to improve profitability. Morgan Stanley's revised forecast reflects this urgency: the earlier 10% projection now seems conservative in light of actual announcements.
Yet, the forecast is not without uncertainties. The conversion of productivity gains into actual headcount cuts depends on how banks balance efficiency against reputational risk. Large-scale layoffs in Europe can provoke political backlash and labour unrest. In Germany, for instance, banks must negotiate with powerful works councils that have the legal right to object to redundancies. In France, labour laws mandate lengthy consultation processes and severance packages. These constraints could slow the pace of job losses, pushing the actual figure closer to 10-15% rather than the full 20%.
Furthermore, the nature of AI adoption is not uniform across all banking functions. While back-office and compliance are most vulnerable, front-office roles that require client relationships and complex negotiation may be less affected. Some banks are even hiring more data scientists and AI engineers to manage the new technology, partially offseting the cuts. The net effect is a recomposition of skills rather than a simple reduction. However, the demand for new roles is smaller in number than the traditional positions being eliminated, resulting in a net decline.
The wider European labour market could feel the impact acutely. Banking employs approximately 2 million people in the European Union, with additional workers in related financial services. A loss of 400,000 jobs would represent a 20% contraction, potentially triggering ripple effects in real estate, retail, and services concentrated in financial hubs like London, Frankfurt, Paris, Zurich, and Amsterdam. Policymakers are already discussing reskilling programmes and social safety nets, but the scale of change may overwhelm existing systems.
The technology itself is evolving rapidly. Generative AI models like GPT-5 and Mistral are becoming more capable, and their integration into banking software is accelerating. Cloud computing and APIs allow banks to embed AI into legacy systems without massive infrastructure overhauls. The cost of AI deployment is also falling, making it accessible to smaller banks as well. This democratisation means that even mid-sized lenders could adopt AI aggressively, increasing the total job displacement.
Morgan Stanley's analysis also highlights a shift in the narrative from 'cost-cutting' to 'productivity-led restructuring'. Banks are framing these moves as necessary for competitiveness, arguing that without AI, they would lose market share to more efficient rivals, including fintech companies and big tech firms entering financial services. This framing may help mitigate political backlash, but it does not change the human cost.
Looking ahead, the next few years will be critical. By 2027, many of the announced cuts will be implemented or underway. If the regulatory environment remains supportive and productivity gains continue to materialise, the 20% figure could become reality. If not, the final number may settle somewhere in between. But the direction is clear: European banking will be a meaningfully smaller-by-headcount industry in 2030 than it is today. If the cuts will hit 200,000 jobs or 400,000 will define how disruptive the transition feels to the wider European labour market.
Source: TNW | Eu News