The Three Percent Revolution

European banks plan to cut 200,000 jobs by 2030 as AI takes hold. The Financial Times reports Morgan Stanley's analysis of 35 major lenders expecting 10% workforce reduction. Société Générale's CEO has declared that "nothing is sacred." The language speaks of disruption while the numbers suggest otherwise.

Two hundred thousand sounds like a reckoning. But ten percent over five years? That just sounds like standard attrition.

Does ten percent over five years actually change anything?

Ten percent over five years works out to 2% annually. Automation has always worked this way in large institutions: managed decline in specific functions, absorbed through retirement and resignation. The pattern is so established it barely registers as news, until someone adds the phrase AI to the announcement.

A revolution that fits neatly into the natural cycle of staff retirement may be better characterised as incremental automation.

And 10% is the optimistic projection. Internal efficiency targets rarely survive implementation. Regulatory friction, legacy systems, and institutional resistance take their share. A 10% headline becomes 5-6% in practice. Perhaps 3%.

The Bloomberg Intelligence survey of 93 global bank CIOs puts it lower still. Across major institutions, the projected average net workforce reduction over three to five years is 3%. The cuts concentrate in back office and middle office functions.

Three percent is not a forecast of disruption. It is a forecast of managed decline in support functions, timed to coincide with natural attrition. Banks have done this before. What AI is actually doing inside banks

Banks are deploying AI to answer customer queries, summarise documents, assist with code, and flag fraud. These are process improvements. They do not change what banks sell, how they fund themselves, or how they take risk.

The Evident AI Tracker confirms this across the sector. Current deployments cluster around operational efficiency. Product innovation and balance sheet restructuring remain untouched.

McKinsey's March 2025 assessment was equally blunt: "Most banks are still experimenting with proofs of concept, such as document summaries or basic emails. Few have identified a path to value or seen substantial returns."

Commonwealth Bank's July reversal

Commonwealth Bank of Australia offered a practical test of these projections in August 2025. The bank had deployed AI voice bots in June, claiming they reduced call volumes by 2,000 per week. Based on this figure, 45 customer service workers were declared redundant and terminated in July. The Finance Sector Union disputed the numbers and took CBA to Fair Work Commission. Call volumes, it emerged, had increased rather than decreased. CBA was forced to draft team leaders onto phones and offer remaining staff overtime.

By August, the bank reversed the terminations, publicly stating that its "initial assessment did not adequately consider all relevant business considerations and this error meant the roles were not redundant." The bank apologised and offered rehiring.

The voice bot may have worked fine. The productivity metrics underlying workforce decisions required recalibration to account for operational complexities. CBA's error was measurable and public. Most such calculations never face a tribunal.

AI as justification, not cause

Cost-cutting programmes, branch closures, and workforce consolidation have been constants in banking for three decades. AI now appears in the announcements. For many of these 200,000 roles, the decision would have looked the same five years ago. The press release would have cited something else. The pattern of reductions is familiar. The tools being deployed are incremental. If AI were driving a different kind of change, the shape of the cuts would be different. It is not.

Even on the banks' own numbers, AI is not coming for traders and dealmakers first. It is coming for back office, middle office, and operations. These are the same functions that get squeezed whenever cost pressure rises.

Investor expectations and the credibility gap

Amazon's Beth Galetti called AI "the most transformative technology we've seen since the Internet," yet the company eliminated 14,000 corporate jobs, just 4% of its corporate workforce. Walmart CEO Doug McMillon claimed "every job we've got is going to" be affected by AI while explicitly maintaining flat headcount of 2.1 million over three years.

Publicly frame AI as transformative because investor pressure demands bold narratives. Privately implement incremental automation because that is what works.

What this argument is not claiming

This is not an argument that AI is trivial. It is an argument that, inside banks, AI is being used the same way every other efficiency tool has been used for thirty years: to shave points off cost.

What the forecasts tell us

The sector has already priced AI into its own projections. The result is 2% a year, absorbed through attrition, concentrated in support functions.

If that is the best the banks can project for themselves, their forecasts are telling us something about the limits of AI in a regulated industry. Strategic planning requires balancing transformative vision with operational realities reflected in current deployment metrics. In other words; the hype is optional. The numbers are not.

References

Bloomberg Intelligence. (2025). Global Bank CIO Survey: AI Workforce Impact Projections.

Boston Consulting Group. (2025). For Banks, the AI Reckoning Is Here. Available at: https://www.bcg.com/publications/2025/for-banks-the-ai-reckoning-has-arrived

Evident Insights. (2025). 167 Ways Banks Use AI. Available at: https://evidentinsights.com/bankingbrief/167-ways-to-use-ai-special-edition/

Financial Times. (2025). Morgan Stanley predicts over 200,000 European banking jobs will be cut by 2030.

McKinsey & Company. (2025). How banks can turn AI's promise into real impact. Available at: https://www.mckinsey.com/industries/financial-services/our-insights/banking-matters/how-banks-can-turn-ais-promise-into-real-impact

MIT News. (2024). What do we know about the economics of AI? Available at: https://news.mit.edu/2024/what-do-we-know-about-economics-ai-1206

Newsweek. (2025). Walmart Celebrates Automation as US Job Cuts Reach Multiyear High. Available at: https://www.newsweek.com/walmart-celebrates-automation-us-job-cuts-reach-multiyear-high-11107369

TrendForce. (2025). 2025 Tech Layoffs: Meta, Amazon, Microsoft and Others Cut Tens of Thousands of Roles. Available at: https://www.trendforce.com/news/2025/12/31/news-2025-tech-layoffs-meta-amazon-microsoft-and-others-cut-tens-of-thousands-of-roles/