Before artificial intelligence takes Wall Street jobs, it is first creating them. Job postings at major US banks that reference AI roles surged 49 percent this year against 2025, to 139,819 listings, according to an analysis by enterprise hiring-data firm Draup provided exclusively to CNBC and published on 2 October 2026. The fastest-growing skill inside that pile is one most job titles did not even name a year ago: agent orchestration — designing teams of AI agents that work in concert on a task — with references up 1,721 percent year over year.
🔍 The Bottom Line
The headline story about AI in finance has been job cuts; the hiring data tells the other half. Banks are not hiring armies of model builders anymore — governance, risk and deployment roles now outnumber the training-and-running-models side in the postings — and the skills commanding premium pay are about wiring AI into real business processes, not building it. That is a career signal, not just a tech trend.
📚 What Is Agent Orchestration?
Think of a small crew rather than one clever assistant. Instead of a single chatbot answering questions, an orchestrated system strings together multiple specialised AI agents — one to inspect raw data, another to analyse a document, a third to check regulatory compliance — while a human decides when oversight is needed. Agent orchestration is the skill of designing that crew: choosing which agents are needed, what each one does, and when a person steps in.
One Skill, 1,721 Percent Growth
Draup, which culls job-post data from public listings and platforms including LinkedIn, tracks skill references across bank postings. Agent orchestration jumped 1,721 percent this year — from 108 references to 1,967 — making it, in the words of Draup CEO Vijay Swaminathan, “arguably the hottest skill on Wall Street.” The adjacent stack moved with it: references to LangGraph, a framework for building multistep agent workflows, rose 679 percent; LlamaIndex, which connects AI applications to data, rose 291 percent; and retrieval-augmented generation, the technique for feeding models company data, climbed 259 percent.
What is striking in the data is where the growth is concentrating. Postings mentioning responsible AI surged 657 percent, AI governance rose 394 percent, and AI risk management rose 359 percent. Governance-related skills now account for more than 16,000 references in the Draup data, nearly twice the roughly 8,400 tied to training, deploying and running models. The boom, in other words, is not only for people who build the models — it is increasingly for people who keep the models safe, compliant and actually working.
New Roles, Real Pay
The hiring is expanding beyond engineers and data scientists to forward-deployed engineers — people who embed AI directly into trading desks, compliance units and back-office operations, and who need domain knowledge as much as technical ability. LinkedIn’s own Talent Connect research showed this year that forward-deployed roles grew 226 percent in a year across the wider market; Draup’s banking numbers suggest finance has joined that shift.
The pay follows the scarcity. Roles tied to generative AI and agents typically pay more than tech roles elsewhere in finance, per Draup: a generative AI manager’s median base salary sits at about $190,000, agentic AI engineers at $177,000, and chief data scientists at $235,158 — before bonuses and equity. Swaminathan notes the soft skills matter as much as the stack: “problem solving, creativity, ability to ask tough questions, being assertive” when it comes to understanding the processes being automated.
For banks, filling the roles is hard enough that they are leaning on internal reskilling — training existing developers and domain experts rather than competing for scarce specialists. JPMorgan CEO Jamie Dimon has spoken of “huge redeployment plans” as AI takes over more work, a program that treats AI primarily as a reassignment engine rather than only a redundancy engine. It is the mirror image of the April pattern we covered, when six big banks cut 15,000 jobs while posting $47 billion in profits: the same AI strategy produces both numbers.
The NZ Reading
New Zealand’s banks and insurers sit a step behind the US majors, but the job boards are already signalling the same redefinition: SEEK NZ’s August report showed AI skills requested in 4 percent of local ads, up 93 percent annually, even as total listings grow steadily. The American data suggests which part of the skill tree to watch: not “learn to code” in the abstract, but the pairing of domain knowledge — compliance, risk, operations — with the ability to configure and oversee AI systems. That pairing is teachable, and it is exactly where internal reskilling programs concentrate. It is also, per the PwC survey showing finance professionals rate AI skill above an MBA, already how the market prices junior talent.
❓ FAQ
What is agent orchestration? Designing teams of specialised AI agents that work together on a task — deciding which agents are needed, what each one does, which tools they use, and when a human overseer must step in.
How fast is demand for agent orchestration growing? References in bank job postings jumped 1,721 percent in 2026, from 108 to 1,967, according to Draup’s analysis for CNBC.
Are banks hiring more AI roles or cutting more jobs? Both are happening in parallel. AI-related postings at major banks rose 49 percent this year even as banks continue restructuring — the roles being created differ from the roles being cut.
What do these AI roles pay? Draup’s banking-sector medians (base pay only): about $190,836 for a generative AI manager, $176,999 for an agentic AI engineer, and $235,158 for a chief data scientist.
What could a New Zealand worker do with this? The demand concentrates on domain-plus-AI pairings — risk, compliance and operations people who can configure and govern AI systems — which are the same skills NZ banks and insurers’ internal reskilling programs target, and they are already the subject of 4 percent of NZ job ads.
🔍 The Bottom Line
On the evidence of the postings, AI’s first big act in finance was not deletion — it was the invention of a job category nobody was hiring for a year ago, attached to five-figure pay premiums and a governance workload twice the size of the modelling one. Whether the agents eventually consume as much work as they create remains unresolved; what the 1,721 percent number settles is where the leverage sits while that question is open.
📰 Sources
- CNBC — How AI is redefining Wall Street jobs — and boosting demand for this new ‘hottest skill’ by 1,721%
- Draup — bank job postings analysis provided exclusively to CNBC (estimates as of September 2026)
- Vijay Swaminathan, CEO of Draup — quotes from CNBC interview
- JPMorgan — Jamie Dimon on AI redeployment plans (February 2026)
- Singularity.Kiwi — LinkedIn AI safety roles data
- Singularity.Kiwi — Wall Street AI layoffs analysis
- Singularity.Kiwi — SEEK NZ job ads and AI skills
- Singularity.Kiwi — PwC survey on AI skill vs MBA in finance
— CJ Murden, editor of Singularity.Kiwi. Former digital technologies teacher, author of AI-focused books. Writing with a New Zealand focus.
