AI Cuts Finance Teams — Operators Who Build Win

TL;DR: Retail brokers are citing AI as the primary reason for headcount cuts, but CFTE co-founder Huy Nguyen Trieu calls this a trap. Firms that use AI only to reduce staff are choosing jobless growth over expansion. The operators who build talent density and deploy AI as a force multiplier — not a cost lever — are the ones who will own the next cycle.

The Cost-Cutting Trap Finance Operators Keep Falling Into

Several retail brokers in 2026 have publicly credited AI adoption as the main driver behind workforce reductions. On the surface, the math looks clean: fewer salaries, same output, better margins ahead of the next earnings call. Trieu, who co-founded CFTE (Centre for Finance, Technology and Entrepreneurship) and authored The AI-fication of Jobs, sees something more corrosive underneath.

“You can’t grow a company just by cutting,” he says. “If you have unlimited resources, you should be thinking about all the amazing things you can build.”

The trap is psychological. Leaders who frame their teams as cost centres to be minimised are making a category error. AI is not a scalpel for trimming headcount. It is, as Trieu frames it, an engine of unlimited cognitive expansion. The companies that treat it only as a scalpel will eventually run out of things to cut, with no new revenue architecture to show for it.

This distinction matters enormously for high-CAC verticals. Whether you are running forex client acquisition at scale or managing compliance-heavy onboarding flows, the ceiling on what you can build is now far higher than it was three years ago. Cutting your way to that ceiling is not a strategy.

Cognitive Power, Not Distribution: Why This AI Wave Is Different

Fintech’s first decade was fundamentally about distribution. Challenger banks, mobile wallets, and low-cost transfer rails made financial services cheaper and more accessible. That wave lowered friction. It did not change how financial services were constructed.

AI operates at a different layer entirely. “It doesn’t help you with accessibility,” Trieu argues. “What it changes is the level of cognitive power. Finance is a world of highly qualified people and intellectual power. Digital Intelligence is hitting that level.”

In practical terms: we moved from basic chatbots to agentic systems capable of planning and executing complex multi-step tasks in roughly three years. Most institutional structures were not designed to absorb that pace. Their org charts, approval chains, and five-year planning cycles assume a rate of change that no longer exists.

For operators running iGaming player acquisition or crypto token launches, the implication is direct. The firms that modelled on the fintech distribution playbook — spend, acquire, retain — now have access to cognitive tools that can compress the strategy, creative, and analytics work that previously required large specialist teams. The question is whether they are using those tools to build or just to cut.

Supercharged Professionals vs. Task Robots

Trieu draws a clear line between two cohorts emerging inside financial services organisations right now.

Supercharged professionals combine deep domain knowledge with the ability to automate their own manual workflows, plus a mindset built for continuous self-directed learning. They can produce in one hour what previously took a full day. They are not afraid of the tools because they have absorbed them into how they work.

Task robots are skilled at repetitive, codified processes — exactly the work that AI handles best. This group is under direct pressure. Learning to use an AI tool is relatively accessible. Becoming a systems thinker who can identify leverage points, question assumptions, and design new workflows is not. Not every task robot makes that transition successfully.

The gap between these two groups is already affecting hiring, team design, and how performance is measured. Operators using performance ad management services should be asking the same question internally: are the people touching their campaigns supercharged or task-dependent? Because the cost of task-dependent thinking is now measurable in lost yield on every dollar spent.

There have been clumsy attempts to force the transition through blunt metrics. Some firms experimented with “tokenmaxing,” evaluating employees based on raw volume of AI tool usage. Employees gamed it immediately. Crude output metrics cannot substitute for genuine systems thinking.

The Junior Bottleneck: A Structural Problem Operators Cannot Ignore

The most structurally damaging effect of this transition may be the hollowing out of junior talent pipelines. Entry-level hiring has become a Catch-22: companies want experience, but the junior roles that built that experience are the first ones being automated away.

A large-scale study tracking online job vacancies found that in the year following mainstream GenAI adoption, junior software developer postings dropped more than 16% relative to senior positions. The same pattern is visible across financial services, legal, and marketing operations.

Trieu is working with Singapore’s Institute of Banking and Finance on a structural fix: train graduates in applied AI and specific financial domain knowledge before they enter the workforce, so they arrive capable of contributing immediately rather than spending 18 months learning the basics that AI now handles. The objective is to compress or bypass the traditional junior ramp entirely.

For operators managing crypto marketing pipelines or legal firm lead generation, the downstream implication is real. The agencies and in-house teams they rely on are facing the same bottleneck. Mid-level talent that would normally have been built up through junior roles over three to five years simply does not exist in the same volume anymore. Planning for that shortage now — rather than when a campaign is mid-flight — is the move.

What This Means for High-CAC Vertical Operators

Forex, iGaming, crypto, and legal are all high-CAC environments where the cost of a bad hire or a slow team compounds fast. The dynamics Trieu describes have specific operational consequences for these verticals.

First, the “stamp culture” — hiring based on credential trails rather than demonstrated capability — is becoming a liability. The right university, the right internship, the eventual managing director track: this framework still exists, but the number of people who can navigate it productively is shrinking. Operators who rely on credential filtering rather than output filtering will pay a premium for talent that is not necessarily more capable.

Second, the five-year plan is dead. Trieu is blunt about this. The underlying technology stack changes every six months. A planning horizon that extends five years is not strategic; it is speculative fiction. Operators who run precision audience targeting know that campaign assumptions from 18 months ago are already outdated. The same is true for team structure, tool selection, and budget allocation.

Third, talent density beats headcount. Trieu’s framing here is actionable: you do not need to transform the entire organisation at once. Start with a small cohort of supercharged operators who influence the rest. A full marketing audit is often the fastest way to identify where cognitive leverage is already available and going unused — which channels, which workflows, and which targeting assumptions are still running on task-robot logic when the tools to upgrade them are already available.

The operators who treat AI as a force multiplier for their best people, rather than a replacement for their cheapest, are building something structurally durable. The ones cutting their way to margin are just delaying a reckoning.

Originally reported by Finance Magnates Executives, June 2026.