The New AI Job Titles, by the Numbers: AI Engineer, Agentic Engineer, and the Roles That Didn't Exist 3 Years Ago

In January 2023, "Generative AI Engineer" was not a job. Lightcast counted 55 postings worldwide that asked for generative AI skills. By May 2025 that same skill bucket showed up in nearly 10,000 monthly postings, and "AI Engineer" had become the single fastest-growing job title in the United States for two years running on LinkedIn's annual ranking. (Lightcast, LinkedIn)

The titles are still in flux. Some of them are real engineering disciplines. Some are rebrands of work that already existed. Some will be gone in 18 months, absorbed into broader buckets. This post is an honest, data-first look at where each title sits today, how fast it's growing, what it pays, and what skills it actually demands, with citations to the source data so you can check the math.

1. The market context: AI postings detached from the rest of tech

The cleanest way to see what is happening is to compare AI-mention job postings against all job postings on the same platform. Indeed's Hiring Lab has been publishing this since 2020, and the two lines diverge sharply starting in late 2023.

At the end of 2025, the Indeed AI Tracker (share of all postings that mention AI or AI-related keywords) hit a new high of 4.2%, up from a 2022 peak of 3.3% that then fell to 1.7% by mid-2023 before ChatGPT-driven demand reversed the trend. Total tech postings on Indeed finished 2025 about 34% below their pre-pandemic baseline, while tech postings mentioning AI were about 45% above it. (Indeed Hiring Lab, CIO Dive)

PwC's 2025 Global AI Jobs Barometer puts the absolute level around 874,000 AI-skill US postings at the 2022 peak, with continued share growth through 2024 even in a softer overall market. (PwC)

2. The generative AI skill bucket grew 180x in 4 years

Lightcast's tracker for postings explicitly requiring generative AI skills (LLMs, prompt engineering, RAG, fine-tuning, agent frameworks) starts from a near-zero base and has compounded fast.

Lightcast also reports that job postings requiring generative AI skills in non-IT roles are up 9x from 2022 to 2024, postings for Generative AI Engineers specifically are up 7x, and postings requiring generative AI skills in other IT roles are up 35x. In 2019, 61% of AI postings were for tech roles; by 2025 that had fallen to 49%, meaning the majority of AI-skill postings now sit outside the IT and computer science career area. (Lightcast)

3. LinkedIn's fastest-growing jobs are AI, two years running

LinkedIn's "Jobs on the Rise" report ranks job titles by growth in member transitions over a 3-year window. AI Engineer took the #1 spot in 2025 and held it in 2026. Four of the top five 2026 roles are AI-related.

Source: LinkedIn Jobs on the Rise 2026 (Jan 2026). Four of the top five roles are AI-related.

LinkedIn reports that AI Engineer postings rose 143% year-over-year in 2025, and that AI/ML postings overall surged 163% from 2024 to 2025, reaching roughly 49,200 positions in the US alone. Between 2023 and 2025 the platform added 639,000 AI-related job postings in the US, 75,000 of which used the AI Engineer title. (LinkedIn, CBS News, HeroHunt)

For AI Engineer specifically, LinkedIn's role profile is now stable: most common skills are LangChain, Retrieval-Augmented Generation (RAG), and PyTorch; most common industries are Technology and Internet, IT Services, and Business Consulting; top metros are San Francisco, New York, and Dallas. (Dice)

4. The titles, by year they emerged

Second Talent's 2026 report on AI hiring tracks the year each new AI title first appeared in meaningful volume on their engineering hiring platform. The list is a useful taxonomy because it draws a clean line between the "first wave" titles (LLM Engineer, Prompt Engineer, AI Automation Engineer) and the second-wave agent titles (AI Agent Engineer, RAG Engineer, LLMOps Engineer).

Source: Second Talent 2026 AI Roles Report.

Second Talent reports that AI Automation Engineer and AI Agent Engineer roles together accounted for 38% of all 2025 placements through their platform. Both roles barely existed in 2023.

5. The standouts: agent and forward-deployed engineering

A handful of titles are growing faster than the AI Engineer umbrella, with year-over-year growth that is genuinely unusual for any job title.

LangChain's State of Agent Engineering 2025 report cited 986% YoY growth in Agent Engineer postings. Forward-Deployed Engineer postings, the role popularized by Palantir and now common at OpenAI, Anthropic, and AI infrastructure companies, grew over 800% in 2025 according to a synthesis of Indeed and LinkedIn data. AI Agent Engineer postings on Indeed grew 10x year-over-year per WSJ data cited by Gergely Orosz, and Second Talent measured 240% growth for the same title in their 2025 placements. (Product Leaders Day India, HeroHunt, Ivan Turkovic)

The phrase "agentic engineering" itself is starting to displace "software engineering" at the frontier. Boris Cherny, who works on Claude Code at Anthropic, posted in May 2026 that "programming via LLM agents is increasingly becoming a default workflow for professionals" and proposed agentic engineering as the new name. His prediction that the job title "software engineer" will begin to disappear in 2026, absorbed into more hybrid roles, has been widely cited. (AgenticMSP)

6. The skill stack: what an AI Engineer posting actually asks for

Two independent analyses, one of 3,449 postings and one of 31 hand-picked postings at top AI companies, converge on the same picture: Python plus LLMs is table stakes, and the LLM application stack (RAG, LangChain, agents, evals) has crossed from differentiator to baseline.

The Heavybit analysis (top AI companies) shows two big divergences from the broad market: evaluations appear in 61% of postings at top AI companies versus only 8% in the broader market, and agentic workflow mentions appear in 71% of top-tier postings. PyTorch is actually lower at top AI companies (16% vs 23%) because the role isn't training models, it's building applications that call them. (Dev.to, Heavybit)

Quoting the Heavybit analysis directly: "AI Engineers are applied AI engineers, software engineers familiar with the tooling around LLM apps and how to wrangle accuracy out of a stochastic compute stack."

7. The other big one: Chief AI Officer

The CAIO title has gone from novelty in 2023 to near-standard at large enterprises in 2026. IBM's Institute for Business Value, in research debuted at Think 2026, documented one of the fastest C-suite adoption curves in modern corporate history.

CAIO job postings are up 340% since 2023, and 57% of CAIOs in 2026 were promoted from internal talent pools rather than hired externally. Among FTSE 100 companies, nearly 48% now have a CAIO or equivalent, with 65% of those appointments made in the past two years and 42% since January 2024. The role increasingly reports directly to the CEO. (Christian & Timbers, Foundry State of the CIO 2025)

The "Head of AI" role (one level below CAIO) is growing in parallel. LinkedIn's January 2026 Davos release reports double-digit increases in companies with a Head of AI position across Australia (+32%), Canada (+31%), India (+30%), Germany (+30%), the UK (+30%), and the US (+28%) in a single year. (LinkedIn)

8. Salaries: the premium is real, and growing at the staff level

Levels.fyi's 2025 compensation report shows that AI-focused Software Engineers earn $245,000 median in the US, with the gap over non-AI engineers widening at the senior end of the career ladder.

The entry-level premium is actually shrinking (10.7% in 2024 down to 6.2% in 2025), because the market is leveling out as more juniors enter with relevant skills. The staff-level premium is widening (15.8% to 18.7%), because senior AI talent is genuinely scarce. (Levels.fyi)

Specific titles, with their published median total compensation:

Sources: Levels.fyi AI Engineer page, Levels.fyi Google AI Engineer, Glassdoor via Coursera (Prompt Engineer), Glassdoor Agentic AI Engineer via Ivan Turkovic, Christian & Timbers (CAIO).

The Levels.fyi compensation history for AI engineers tells a slightly different story than the simple "up and to the right" narrative: median total comp peaked at $295,000 in March 2024, fell to $228,500 by January 2025, and has rebounded to $260,000-$269,000 by the end of 2025. The 2024 peak was driven by large Meta packages; the 2025 floor coincided with broader tech market weakness. (Levels.fyi)

For Forward-Deployed Engineers, the comp band sits at $280K-$450K+ total, and Staff/Principal-level Agentic AI Engineers at leading agent shops reach $350K-$500K TC. (Product Leaders Day India, Ivan Turkovic)

9. The geographic concentration is intense

LinkedIn's role profiles repeatedly surface the same three to five metros across every AI title. The University of Maryland / LinkUp AI Job Tracker confirms the same pattern at the state level: California, Washington, and Texas dominate. (Axios)

San Francisco employs the largest share of AI engineers and researchers in the US. Boston-Cambridge leads in biotech R&D that is itself becoming more AI-centric. New York is the financial-services AI hub and the second-largest market overall. Seattle ranks #2 nationally for raw AI posting volume, with 1,472 postings tracked in January 2025. (R&D World)

Remote and hybrid options for AI Engineer roles sit at roughly 62.8% combined (35.5% remote, 27.3% hybrid) per LinkedIn's role profile, higher than the average tech role. (R&D World)

10. Sector mix: AI hiring is leaking out of tech

When generative AI hiring first took off in 2023, it was almost entirely concentrated in software companies. By the end of 2025, Indeed Hiring Lab reported that 45% of data and analytics postings mention AI, but the more interesting growth is in non-tech sectors.

The marketing share nearly doubled in 2025 (from 8.4% to 14.9%) and the HR share doubled (4.4% to 8.8%). PwC's barometer notes that the Health and Social sector's overall AI posting share rose from 9.6% in 2012 to 14.7% in 2024, and that healthcare was the single largest creator of AI jobs in 2025, generating more than 640,000 positions linked to diagnostics, predictive analytics, and virtual patient support. (Indeed Hiring Lab, PwC, HeroHunt)

11. The wage premium is showing up in the data, not just the anecdotes

PwC's 2025 Global AI Jobs Barometer found that workers with AI skills command a 56% wage premium on average, and that jobs requiring AI skills continue to grow faster than overall job postings. That premium has held steady as the broader market has softened. (PwC)

A consortium-led 2025 industry survey reported that 78% of IT job roles now list AI expertise as a requirement. That number is up sharply from prior years and is the clearest signal that AI is no longer a specialty skill set inside tech, it is becoming a baseline. (IntuitionLabs synthesis)

12. Which titles will survive, and which won't

The honest read of the data: the AI Engineer umbrella is going to absorb most of the sibling titles over the next 18 months, with two or three specialist titles surviving as genuine sub-disciplines.

Ivan Turkovic's 2026 AI Job Title Reference Guide makes the strongest case for the consolidation. His read of the market:

• AI Engineer is becoming the default umbrella. "Will absorb most sibling titles over the next 18 months."
• GenAI Engineer peaked as a distinct title in late 2024 and is consolidating into AI Engineer.
• LLM Engineer is AI Engineer plus fine-tuning and deeper eval frameworks. Still a real specialty.
• Prompt Engineer is declining as a standalone title, being absorbed into AI Engineer and Context Engineer roles. Comp range still wide ($63K to $216K depending on seniority).
• Context Engineer is a genuinely new specialty: owns retrieval, memory, tool definitions, system prompts, and metadata governance. Cognizant announced 1,000 Context Engineer hires in August 2025. Likely survives.
• Agentic AI Engineer / Agent Engineer is real work but distinguishing from AI Engineer is fuzzy. Likely settles into "AI Engineer (Agents)" as a discipline tag.
• RAG Engineer is niche and fading. Being absorbed into AI, LLM, and Context Engineer titles through 2026.

Source: Ivan Turkovic, The AI Job Title Reference Guide 2026

The big-picture forecast: WEF estimates 12 million new AI-related jobs globally by the end of 2026, even as 92 million traditional roles face disruption. The World Economic Forum has documented over 1.3 million AI-enabled jobs already created globally, with AI Engineers, Forward-Deployed Engineers, and Data Annotators leading the count. (LinkedIn Davos 2026, HeroHunt)

13. What this means if you're job-hunting

Three things stand out from the data once you stop looking at headline growth rates and start looking at the underlying postings:

1. Python plus LLMs is the floor, not the differentiator. They show up in 66-77% of postings depending on the analysis. If you don't have both, you're not in the pool. The skills that actually move offers are evaluations (61% of top-tier postings, 8% of broad market) and shipped agentic workflows.

2. Title inflation is real, but the work is genuinely new. "Agentic AI Engineer" might be a marketing term, but the day-to-day work (multi-step planning loops, tool-use frameworks, agent memory, observability) is meaningfully different from what an ML engineer was doing in 2022. Don't dismiss the titles, but optimize your search for the work, not the label.

3. The geographic concentration is brutal. If you're not in or willing to relocate to SF/NYC/Seattle/Boston, you're competing for the ~30% of postings that are remote-friendly. That subset is growing, but it's the most contested subset of the most contested job category in tech.

If you're applying to AI roles at any scale, the volume problem is real. The strongest signal in the data is that roles are growing faster than candidate supply: ManpowerGroup noted a 16% increase in AI-related postings in a single three-month stretch in 2025, while overall tech hiring fell 27% in the same window. (IntuitionLabs) That gap is your opportunity, but only if you can apply at the volume the market expects. That's where automated applying changes the math.

Apply to every new AI role the day it goes live

Here is the brutal math from the data above. AI Engineer postings grew 143% in a single year. The roles cluster in the same handful of metros, so 75,000+ new postings between 2023 and 2025 turn into a few hundred high-fit openings for any one candidate. Most of them get filled within 14 days. Job seekers who win in this market are not the ones with the best resume, they are the ones whose resume lands in the first 50 applications.

Job Agent solves this:

• Daily scrape of 600+ direct company job boards the moment a new role posts. We don't wait for aggregators.
• Match scoring against your profile so we apply only to roles that fit your title, location, seniority, and salary band, not blast-spray.
• AI fills the application including the open-ended "tell us about a time..." questions, using your resume + answer history. ATS-aware, not autofill.
• Email confirmation + interview tracking by scanning your Gmail so you see the funnel, not just the submissions.
• Up to 1,500 applications a month on the Ultra plan. Starter is $19/mo for 25 apps, Pro $49 for 100, Growth $99 for 250.

Try Job Agent free → www.jobagent.sh 3 free applications/month. No credit card. Live in 5 minutes once you upload a resume.

If you are an AI Engineer, Agentic Engineer, ML Engineer, or anywhere in the 10-title cluster above, this is the highest-leverage move you can make in a market this concentrated.

Sources

Primary data sources used throughout this piece:

• Indeed Hiring Lab - January 2026 US Labor Market Update, March 2024 US Labor Market Update, September 2023 AI Jobs Tracker
• Lightcast - The Generative AI Job Market: 2025 Data Insights
• LinkedIn - Davos 2026 Labor Market Report, Jobs on the Rise 2026 (via Dice), Jobs on the Rise 2025 (via Axios)
• Levels.fyi - AI Engineer Compensation Trends Q3 2025, Google AI Engineer salaries
• PwC - 2025 Global AI Jobs Barometer (US)
• IBM Institute for Business Value (via Christian & Timbers) - How to Hire a Chief AI Officer in 2026, Why Boards Are Racing to Appoint Chief AI Officers in 2025
• Second Talent - 2026 AI Roles Report
• Dev.to - 3,449-Posting Analysis of AI Engineer Skills (2026)
• Heavybit - So You Want to Be an AI Engineer
• Ivan Turkovic - The AI Job Title Reference Guide 2026
• CIO Dive / HR Dive - AI jobs resistant to recent tech hiring slump
• R&D World - 2025 LinkedIn report: AI engineers lead R&D/STEM growth
• HeroHunt - Fastest Growing AI Roles in 2026: Data and Rankings