Discussing the Wave of Fintech Layoffs with AI Optimization Trend
The financial sector is currently experiencing one of the most significant waves of optimization in recent years — and behind it is not a crisis, but a technological shift. We have covered the most prominent cases on multiple occasions: BILL cutting 30% of its workforce, PayPal planning to eliminate 20% of employees (~4,700 people) over two to three years, Block laying off 40% of its staff, and Coinbase — 14%. All of them cite the same reason: AI implementation allows them to do the same work with fewer people. Companies want to earn more, spend less, and move faster. Some fintech have introduced revenue-per-employee benchmarks, and the logic follows: the fewer employees a company has and the higher its profit, the more efficiently it operates.
Expert and employee opinions are divided. Some believe this is hype and companies will soon start hiring people back — as was the case with Klarna. Others are convinced the changes are irreversible and logical. It all looks like a sharp, stressful leap, yet businesses are reporting strong profits and have simply decided not to drag things out — restructuring their processes in one decisive move.
Monoup has been tracking these trends for some time, and today we break down what is actually going on.
Why Now, And Who is AI Replacing First
It will come as no surprise that mass layoffs started with frontline positions that have historically employed large numbers of people. These roles existed because they were necessary for scaling and processing growing volumes of data — something that simply could not be done with a small headcount. Fintech grew rapidly up to a certain point, after which slowdown and a period of relative calm typically follow. When growth decelerated and AI tools caught up with real-world workflows, it became clear that a significant portion of headcount had been a buffer for volume, not for actual complexity. Large companies frequently hired a reserve, operating on the assumption that you can always find a task for people and that someone would be available to cover colleagues during vacations. The result was an accumulation of personnel costs that no longer deliver the same return.
Customer support is the most obvious and fastest candidate for replacement by agents. The work in these departments is structured, scenarios repeat, and responses are trainable. PayPal explicitly names customer service and merchant support as first in line for cuts. Klarna replaced hundreds of operators with AI chatbots as early as 2024–2025 — and in the initial phase, this genuinely delivered savings. By the end of 2025, however, the company quietly began rehiring, as the quality of complex query handling had deteriorated. Block followed the same trajectory — after mass layoffs, it began bringing back former employees in engineering and other key functions by March 2026. The numbers rehired never reached original levels, but the direction is clear: companies overestimated their readiness at that point, without abandoning the approach altogether. The future of customer support looks like this: 90% of cases resolved by AI assistants, with only the most complex 10% handled by real people. Support specialist remains an existing role — but hiring volumes are shrinking dramatically, and the department will likely cease to be the largest in most companies.
Fraud detection and monitoring specialists, along with AML professionals, are the second group facing optimization — though this receives far less attention in the industry. Transaction monitoring is, at its core, rules plus patterns, and AI handles this at scale far better than a team of analysts. Coinbase states that monitoring and fraud prevention are priority areas for automation right now — unsurprisingly, since for crypto exchanges these tasks are crucial, time-consuming, slow down processes, and currently come at high cost. This does not mean regulatory expertise will disappear, but routine monitoring as a standalone function is clearly contracting.
Middle management is the least obvious but most telling group. Block and Coinbase are deliberately removing the coordination layer of middle managers — people who do not build the product but pass information between those who do. Their primary role is to be strong communicators and keep an eye on all processes, which is precisely why most companies hire middle management in the first place. But Coinbase is moving to a “player-coach” model, where a manager retains their role only if they are also a strong individual contributor. Also, the company is creating |AI-native pods” — teams of one person managing agents in place of an entire department of engineers, designers, and product managers. “Layers slow things down,” says Coinbase CEO Brian Armstrong — and this is an operational decision he is actively executing. This approach could well become a template for other companies.
Which Fintech Roles Are at Risk
Beyond those already mentioned, other roles are also likely to face AI-driven replacement in the near future. Junior analysts are among them — those whose work consists of building dashboards and running routine SQL queries. The value of an analyst is shifting from data extraction toward interpretation and decision-making, which analysts in this category typically do not do. In large companies, decisions rest with domain specialists, while analysts supply the data. The impact could ripple in both directions: junior analysts lose their jobs, while professionals in other fields are pushed to learn SQL themselves and build their own reporting. AI-powered tools will make this substantially easier and faster, eliminating the need to wait weeks for a report from the analytics department.
Recruitment is another function likely to face optimization. Tasks such as CV screening, interview scheduling, and first-round interviews lend themselves well to automation. Block has already reduced its recruiting headcount, signaling that the ground is being tested. Some candidates in the market are already reporting experiences of interviewing first with an AI agent before speaking with a hiring manager. Recruiters at early stages typically work from a standard list of questions and assess candidate culture fit. The former is clearly something agents can handle; the latter remains less certain. What can be said with confidence is that the first attempts to automate hiring are already underway — not only through ATS systems, but through purpose-built interviewing bots.
Some companies also note that junior developers may be entering a zone of risk, as non-technical employees working alongside AI agents are already handling simple tasks. The quality of such output is not yet high, but the direction has been set and specialized agents for repeatable, baseline tasks are actively being developed.
Will There Be New Roles to Count On?
Looking at all of this data, a natural question arises: where does this optimization lead? Will everyone be laid off, or are there other vectors to look at? There are genuine reasons for concern — but there are also opportunities. The new trend does not necessarily mean fewer people in fintech overall; it means a need for fundamentally different people.
Demand is already forming for specialists who can work with AI tools within a specific professional context — not people who simply use ChatGPT, but compliance experts who configure and validate monitoring models, or analysts who interpret what AI cannot explain on its own.
New hybrid roles are emerging — AI Operations specialists who oversee agent quality and manage exceptions. This is essentially what the Klarna case teaches: the first wave of replacement revealed where AI falls short, and those gaps now need to be filled by people with a different profile. The work overall is becoming more complex. High-tech sectors are no longer looking for people who will simply “do a good job” and execute repetitive tasks at volume. They need multifunctional strategists with cross-domain skills. People like this used to be valued exceptions; they are now a baseline expectation for working in big tech.
Specialists in at-risk groups should already be thinking about how to increase their professional value and where they can integrate AI into their existing knowledge. Large data sets, logic, and direct pattern-matching are AI's strengths — but it has clear weaknesses too, and those are precisely where people should be focusing their development.
