Which Industries Are Winning the AI Talent War in 2026?
The AI hiring conversation is dominated by the frontier labs — OpenAI, Anthropic, Mistral. Every week there’s new funding, a new benchmark, and a new surge of applications to a small number of hiring pipelines.
The actual hiring data tells a more distributed story.
HireIndex’s index of 1,718 open AI/ML roles across 784 companies (week of May 12, 2026) places a European AI lab at the top, an enterprise data platform at number two, and a consumer social network in the top 10. A financial services company that rarely appears in AI coverage is hiring more LLM Engineers than any frontier lab in the index. The distribution is not what most candidates are optimizing for.
Which companies are posting the most AI/ML roles right now?
Here’s the top 10 by open role count as of mid-May 2026:
| Company | Open AI/ML Roles | Sector |
|---|---|---|
| Mistral AI | 58 | AI Lab |
| Databricks | 53 | Enterprise Data |
| Scale AI | 51 | AI Services |
| Spotify | 44 | Consumer Tech |
| Amazon | 42 | Cloud / Enterprise |
| 30 | Consumer Tech | |
| Cresta | 29 | Enterprise AI |
| Writer | 25 | Enterprise AI |
| Applied Intuition | 21 | Autonomous Systems |
| Perplexity | 20 | AI Search |
The top 10 is not what most candidates are targeting. Three entries are broadly enterprise software. Two are consumer tech companies. One builds autonomous vehicle simulation. Only Mistral AI and Scale AI fit the frontier-lab archetype that absorbs the majority of public attention.
The top 15 companies combined account for roughly 28% of all open roles in the index. The remaining 72% is distributed across 769 other companies — the majority of the AI/ML job market sitting below the threshold that shows up in any given LinkedIn feed.
What does the sector breakdown actually reveal?
AI labs dominate in volume per company, but fewer companies occupy the category. Mistral AI at 58 open roles leads the entire US-dominated index from Paris — a signal about European AI investment that tends to get underweighted in US-centric coverage. Scale AI at 51 reflects a hybrid position: it is both an AI company and a data labeling and evaluation layer for other AI companies, so its hiring expands proportionally to the overall market.
OpenAI is worth noting as an important contextual data point here. OpenAI holds more than 65 open AI/ML roles in San Francisco alone — accounting for over 50% of that city’s entire tracked AI/ML market. Remove OpenAI from San Francisco’s count and the city looks roughly equivalent in tracked volume to Toronto (81 roles). That concentration is unusual and creates a real candidate funnel risk: a single company drawing the majority of applications into a single city.
Enterprise data and software is the quietly dominant sector. Databricks (53 roles) builds the infrastructure that ML teams run on: data lakes, model training pipelines, feature stores, and the operational layer for running models at scale. Cresta (29) and Writer (25) sit one layer up — applying current-generation LLMs to enterprise workflows in sales intelligence and content generation, respectively. These aren’t research companies building the next frontier model. They are companies deploying existing frontier models for paying enterprise customers. The engineering problems are production-grade: latency targets, reliability requirements, and inference cost at scale.
Consumer tech generates sustained hiring at meaningful volume without the startup risk profile. Spotify (44 roles) runs ML systems that touch hundreds of millions of users: music and podcast recommendation, content understanding, ad targeting at scale. Reddit (30 roles) hires Machine Learning Engineers for recommendation, trust and safety, and search — not AI products in the product-launch sense, but the AI infrastructure keeping a platform with 100M+ daily active users functional. These roles tend to be under-applied relative to their comp and impact, because “ML at a social network” sounds less compelling than “AI at a lab” — even when the engineering problems are harder.
Financial services is the most counterintuitive entrant in the sector picture. Capital One doesn’t make the top 10 in overall role count during this specific data pull, but leads the entire index in one specific category: LLM Engineers. Capital One has 13 open LLM Engineer positions — more than any frontier AI lab tracked in the index. Their New York office holds 47 open AI/ML positions, making Capital One the largest single AI employer in the city — ahead of Spotify (26 NY roles) and every AI lab with a New York presence. JPMorganChase is building a parallel presence in London, where 5 tracked AI roles represent a foothold that will likely grow.
The work at these firms is not theoretical. It’s LLM deployment in production systems where reliability is a regulatory requirement, not an aspiration. The candidate who takes this role seriously is competing against a much smaller applicant pool than an equivalent-seniority applicant to OpenAI or Anthropic — for a role that may involve equivalent engineering complexity.
How does geography interact with the sector breakdown?
The sector distribution is not uniform across cities. San Francisco skews heavily toward AI labs: OpenAI’s dominance (>50% of the city’s tracked roles) makes it the most lab-concentrated market in the index. Candidates targeting SF are effectively targeting a single company’s hiring pipeline for more than half of their options.
New York presents the opposite profile. Its AI hiring is led by financial services and consumer tech — Capital One (47) and Spotify (26) at the top — rather than labs. This is structural: the financial institutions and media companies headquartered in New York have deep data science infrastructure, and the current AI wave is layering LLM capabilities onto that base.
London (173 tracked roles, the largest non-remote market in the index) is more diversified. Spotify (11 roles) and Writer (8) lead, followed by JPMorganChase (5) and Cohere (3), with a long tail across financial services and consulting. No single company or sector captures a majority of the market the way OpenAI does in San Francisco.
What does this mean for job seekers targeting AI roles?
Three sectors account for the majority of substantive AI hiring outside the frontier labs: enterprise software, consumer tech, and financial services. Each offers a distinct tradeoff:
Enterprise software (Databricks, Writer, Cresta): production-grade AI problems, strong compensation, less external brand recognition in AI circles, more predictable risk profile than early-stage startups.
Consumer tech (Spotify, Reddit): high-scale ML infrastructure, genuinely hard engineering problems, recognizable consumer brands, equity structures tied to mature companies rather than frontier AI valuations.
Financial services (Capital One, JPMorgan): serious LLM deployment under regulatory constraints, premium compensation at senior levels, lower equity upside than labs, significantly less applicant volume for comparable roles.
Frontier labs remain the highest-ceiling option for research-oriented engineers. But the data shows that ceiling is narrow: a small number of companies drawing a disproportionate share of applicant attention for a finite number of seats.
The 72% of the market outside the top 15 companies is broader and less picked-over than the lab application funnel that dominates AI job search strategy in 2026. HireIndex tracks all of it — across titles, cities, and skills — updated weekly.