FLUENCY DIGITAL INSIGHTS • WINTER 2025

AI Hiring: A Data-Driven View

Prompt engineering drew early attention, but the data highlights a different set of roles.

The Reality

Prompt Engineer ≠ The Future

Researchers scraped 20,662 AI job postings from LinkedIn across 60 countries. AI job postings from LinkedIn across 60 countries.

How many do you think were for prompt engineers?

↓ Compare to other roles

Source: arXiv:2506.00058 — "Prompt Engineer: Analyzing Skill Requirements in the AI Job Market" (LinkedIn job postings scraped April 7, 2025)
The takeaway: "Prompt engineer" was never a job category the market supported. It's more like a transitional label, similar to "computer operator" in 1985. Eventually, we surmise, everyone will use AI the way everyone uses computers now. The jobs that are emerging are building the systems, not using them as consumers (which everyone will likely be).
The Split

Two Markets, Two Stories

Here's the twist: AI specialist roles (ML engineers, LLM engineers, AI researchers) are still growing. General software hiring is down significantly—the exact causes are debated (interest rates? AI productivity? post-ZIRP correction?), but the divergence is real. (ML engineers, LLM engineers, AI researchers) are still growing. General software hiring is down significantly—the exact causes are debated (interest rates? AI productivity? post-ZIRP correction?), but the divergence is real.

+40%
ML engineer postings YoY
-14%
Overall tech postings YoY
Sources: Bloomberry (ML growth), Dice Tech Job Report (tech decline), KDnuggets (market analysis)
The numbers: Q2 productivity rose 3.3% while hours worked grew just 1.1%. General software postings dropped 25% month-over-month in September. Meanwhile, ML engineer salaries hit 8K average. Many factors are at play, but the split highlights a trend. Q2 productivity rose 3.3% while hours worked grew just 1.1%. General software postings dropped 25% month-over-month in September. Meanwhile, ML engineer salaries hit 8K average. Many factors are at play, but the split highlights a trend.
The Adoption

88% of Companies Now Use AI

McKinsey's 2025 Global Survey shows AI adoption jumped from ~50% (where it hovered for years) to 88% in just two years.

55%
2023
72%
2024
88%
2025
23% already scaling AI agents in production
39% experimenting with AI agents
~6% qualify as "high performers" with enterprise-wide AI impact
Source: McKinsey & Company — "The State of AI in 2025"
The gap: Nearly everyone is using AI, but almost nobody has figured it out. Only ~6% qualify as high performers seeing significant EBIT impact. Nearly everyone is using AI, but almost nobody has figured it out. Only ~6% qualify as high performers seeing significant EBIT impact.
The Risk

Who's Getting Displaced?

Stanford researchers analyzed millions of ADP payroll records—the largest real-time study of AI's labor impact. The finding that made headlines: young workers in AI-exposed jobs are seeing employment decline while older workers in the same roles are growing. while older workers in the same roles are growing.

Source: Stanford Digital Economy Lab — "Canaries in the Coal Mine?" (Brynjolfsson, Chandar, Chen — ADP payroll data, Aug 2025)
The pattern: Erik Brynjolfsson offered one possible explanation: "Older workers have tacit knowledge—tricks of the trade learned from experience that may never be written down anywhere. They have knowledge that's not in the LLMs, so they're not being replaced by them." Other factors could include older workers migrating to fractional work or early retirement—the full picture is still emerging.
Check Your AI Exposure

Based on Stanford's analysis of millions of payroll records. AI will touch every role—the question is how the transition plays out for yours.

--
--

--
Employment trend
--
Experience buffer
The Skills

What AI Roles Require

The jobs are on the creator side—building, deploying, and maintaining AI systems. Machine learning engineer salaries jumped 53% in 15 months while general software engineer pay crept up only 4%.—building, deploying, and maintaining AI systems. Machine learning engineer salaries jumped 53% in 15 months while general software engineer pay crept up only 4%.

75%+ want domain specialists, not generalists
40% ML engineer projected growth (WEF 2023-2027)
Source: 365 Data Science — Analysis of 1,000+ AI engineer postings (2025)
Hiring insight: The roles that matter: MLOps, model deployment, LLM fine-tuning, production systems. The roles that matter: MLOps, model deployment, LLM fine-tuning, production systems.

Hiring AI talent? Looking for your next role?

Fluency Digital specializes in placing senior AI talent at companies building real systems—not hype. Let's talk about what the market actually looks like.

Get in Touch →
About this data: All statistics are from peer-reviewed research, published reports, or primary data sources linked throughout. The Stanford/ADP study represents the largest real-time analysis of AI's employment impact using actual payroll data from millions of workers. LinkedIn job posting data comes from academic research analyzing 20,000+ listings. We update this page as new data becomes available. All statistics are from peer-reviewed research, published reports, or primary data sources linked throughout. The Stanford/ADP study represents the largest real-time analysis of AI's employment impact using actual payroll data from millions of workers. LinkedIn job posting data comes from academic research analyzing 20,000+ listings. We update this page as new data becomes available.