The AI engineer title barely existed five years ago and now sits at the top of LinkedIn’s list of the fastest-growing jobs in the United States. The role sits between software engineering and machine learning: building applications on top of foundation models, wiring them to company data, and keeping them working once real users arrive. This guide covers what the job involves, what the hiring and pay data actually shows, and the skills that keep appearing in the postings.
Updated September 2026. Pay and hiring figures are from the sources named and change quickly.
What the job involves day to day
Most of the work is software engineering with a probabilistic component bolted on. An AI engineer takes a model someone else trained, connects it to the systems and documents a business already has, and turns the result into something that responds reliably, cheaply and safely. The model is rarely the hard part. The retrieval, the evaluation, the cost control and the failure handling are.
- Building on foundation models through provider APIs and SDKs, rather than training large models from scratch.
- Retrieval and data plumbing, so the model answers from a company’s own documents instead of its training data.
- Evaluation, which means deciding what a good answer looks like and measuring it repeatedly, because there is no test suite that simply passes.
- Guardrails and monitoring, covering cost, latency, abuse and the security problems that arrive with untrusted text.
That is different from a research role. According to LinkedIn’s 2026 Jobs on the Rise list for the United States, reported in January 2026, AI and machine learning researcher is a separate entry further down the ranking. The two jobs are often confused by people writing the adverts as well as the people answering them.
How fast the AI engineer market is really growing
LinkedIn’s 2026 Jobs on the Rise list ranks the fastest-growing roles in the United States over the previous three years, and AI engineer came first. Four of its top five entries were tied to AI: AI engineer, AI consultant and strategist, AI and machine learning researcher, and data annotator, followed by data centre technician. LinkedIn listed the most common skills for the role as LangChain, retrieval-augmented generation and PyTorch, and named San Francisco, New York City and Dallas as the cities with the highest concentration of these jobs.
Indeed Hiring Lab gives the wider context, and it is less euphoric. On 31 December 2025 the share of US job postings mentioning AI reached 4.2 percent, its highest recorded level. Those postings stood 134 percent above their February 2020 baseline, while total US postings were only 6 percent above it and the overall Indeed Job Postings Index was down 5.2 percent year on year. AI hiring is growing inside a shrinking market, which is a different story from a boom everywhere.
The spread is also widening. Indeed Hiring Lab counted 822 distinct AI-touched job titles in the United States in the first quarter of 2026, 8.3 percent of all titles, up from 264 titles in the first quarter of 2022. Non-tech occupations accounted for 63 percent of them. In Europe the same measure was smaller: Germany had 288 AI-touched titles in early 2026, the United Kingdom 160.
7 essential skills for an AI engineer
These are drawn from the skills LinkedIn lists for the role and from the published objectives of the vendor exams aimed at it, not from opinion.
- Production Python and ordinary software discipline. Version control, tests, code review and deployment. Microsoft expects a data science background with Python and entry-level DevOps knowledge for its AI operations certification.
- Working with model APIs and SDKs. GitHub Octoverse 2025 recorded 1.13 million public repositories importing large language model SDKs by August 2025, up 178 percent year on year. This is where most of the work now lives.
- Retrieval-augmented generation. Chunking, embedding and searching a company’s own content, usually against a vector database.
- Orchestration frameworks. LangChain appears first on LinkedIn’s skill list for the role. The specific framework matters less than understanding what it is doing underneath.
- Model training fundamentals. PyTorch is on the same list. Few people in the job train large models, but fine-tuning, embeddings and evaluation all assume you know what a training loop does.
- Evaluation and observability. Microsoft’s AI operations exam assesses generative AI quality assurance and observability as a named skill area. If you cannot measure a change, you cannot ship one.
- Security and governance. Anything that reads untrusted text is exposed to prompt injection, and anything that takes actions inherits the risks of an AI agent.
Two more things are worth adding once the basics are in place: prompt engineering as a working habit, and MLOps for anything that has to run on a schedule rather than on demand.
What the pay data actually says
This is where care is needed, because most salary figures circulating online come from job boards and recruiters rather than statistical agencies. The US Bureau of Labor Statistics does not publish a separate occupation called AI engineer. Its nearest categories, with 2025 medians for the United States, are these:
- Software developers: median pay $135,980 a year, with employment projected to grow 10 percent from 2025 to 2035.
- Data scientists: median pay $120,230 a year, 275,600 jobs, with employment projected to grow 35 percent from 2025 to 2035.
- Computer and information research scientists: median pay $140,300 a year, 38,600 jobs, growth of 22 percent from 2025 to 2035, and a master’s degree as the typical entry-level education.
Those are US national medians for 2025 and say nothing about pay in Berlin, Bangalore or Lagos. Treat any single global AI engineer salary number as marketing until you can see the country, the sample and the date behind it.
How people actually move into the role
Indeed Hiring Lab found that between May 2025 and May 2026, 71 percent of the growth in US software development postings came from senior roles and 37 percent from jobs with AI in the title. Software development postings were still about 27.5 percent below their pre-pandemic level. In other words the demand is real but it is skewed towards experience, which matches the fact that most people arriving in these jobs transfer in from software engineering or data work rather than starting there.
The practical route is to build and ship something that uses a model against real data, measure it honestly, and be able to explain what broke. A credential can help structure the learning, and our guide to AI certification options sets out what the main exams cost and cover. If you are weighing this against a less technical path, our guide to becoming a product manager covers a common alternative.
Common questions
What does an AI engineer do? They build software on top of existing AI models: connecting models to company data, designing retrieval, evaluating output quality, and running the result in production with monitoring, cost control and security checks.
Is AI engineer the same as machine learning engineer? The titles overlap and some reports group them. In practice an AI engineer usually builds applications on foundation models, while a machine learning engineer is more likely to train and operate models of their own.
How much does an AI engineer earn? There is no official figure for the title. For the United States in 2025 the Bureau of Labor Statistics reported median pay of $135,980 for software developers and $120,230 for data scientists. Pay varies enormously by country.
Do you need a degree to become an AI engineer? Not necessarily. The Bureau of Labor Statistics lists a bachelor’s degree as typical entry-level education for software developers and data scientists, but employers hiring for this role weigh shipped work heavily.
Is it too late to become an AI engineer? Hiring data from LinkedIn and Indeed through early 2026 shows the role still growing and spreading beyond tech companies, though demand is weighted towards people with existing engineering experience.
Sources and further reading
Where the figures and rules above come from, so you can check them:
- 2026 Jobs on the Rise, fastest-growing US roles (January 2026): LinkedIn, as reported by Dice
- AI job postings share and the December 2025 labour market: Indeed Hiring Lab
- AI-touched job titles across the US and Europe (July 2026): Indeed Hiring Lab
- Software development postings and the senior skew (July 2026): Indeed Hiring Lab
- Software developers: pay and outlook, 2025: US Bureau of Labor Statistics
- Data scientists: pay and outlook, 2025: US Bureau of Labor Statistics
- Computer and information research scientists: pay and outlook, 2025: US Bureau of Labor Statistics
- Octoverse 2025, LLM SDK adoption: GitHub
- Machine Learning Operations Engineer Associate, skills assessed: Microsoft Learn
Photo credits: Programmers by Lbronn, CC BY-SA 4.0, via Wikimedia Commons. Programming code by Martin Vorel, CC BY-SA 4.0, via Wikimedia Commons.
Join the discussion