Skip to content
TBSI.

Start here

  • LLM LeaderboardCompare frontier models on intelligence, price and speedTool
  • What is the best superintelligence?The best frontier model for each jobModels
  • Frontier Gap ReportHow far apart the leading labs areReport
  • MethodologyHow we verify sources, data and claimsAbout

Best AI Jobs (2026): Pay, Growth and How to Get Hired

Applied AI engineer, data scientist, research scientist, ML infrastructure, AI product manager and AI policy roles, with pay and outlook from the US Bureau of Labor Statistics and live job postings

By TheBestSuperintelligence Editorial TeamUpdated

◈ Executive SummaryCited · Documentation

What are the best AI jobs in 2026, and is AI a good career?

The best AI jobs for most people in 2026 are applied AI engineer and data scientist. BLS projects data scientist jobs to grow 35% from 2025 to 2035 at a $120,230 median; research scientists earn more ($140,300) but usually need a master's.

◈Key Takeaways (3)
  • •Data scientist is the fastest-growing AI-adjacent occupation in BLS data: 35% projected growth from 2025 to 2035, median pay $120,230.
  • •Research scientist pays the highest BLS median ($140,300) but is a small field of 38,600 jobs that usually needs a master's.
  • •Applied AI engineer is the widest door: it builds on software development, an occupation of about 1.7 million US jobs.
☰On this page (10 sections)
  1. 01.Which AI job should you aim for?
  2. 02.Is AI a good career in 2026?
  3. 03.Applied AI engineer: the widest door in
  4. 04.Data scientist: the fastest-growing occupation
  5. 05.Highest paying AI jobs: research scientist and research engineer
  6. 06.ML infrastructure and MLOps: the engineers who keep models running
  7. 07.AI product manager, safety and policy roles
  8. 08.How to get a job in AI: your first role
  9. 09.Limits of this guide
  10. 10.Sources & Further Reading

Which AI job should you aim for?

The best AI jobs for most people in 2026 are applied AI engineer and data scientist. Both sit on large, fast-growing occupations in US government data, both accept a bachelor's degree, and both reward people who can show working systems rather than credentials. Research scientist pays the highest median, but it is a small field with a higher degree bar.

The numbers come from the US Bureau of Labor Statistics (BLS) Occupational Outlook Handbook, last modified 2026-08-27, which covers May 2025 pay and projections for 2025 to 2035. BLS does not publish separate figures for "machine learning engineer" or "AI product manager" in the profiles we read, so for those roles we show live postings from four employers (Anthropic, OpenAI, Databricks and Scale AI), each labelled with its company and location. A posting range is what one company advertises for one role, not a market average.

Role What you do Pay source Who it suits
Applied AI engineer Build products on top of models BLS software developers: $135,980 median; Anthropic posting $200,000 to $320,000 Software engineers
Data scientist Model data, run experiments BLS: $120,230 median, 35% growth; Databricks posting $123,200 to $218,000 Analysts, stats and maths graduates
Research scientist or engineer Train and study models BLS: $140,300 median, 22% growth; OpenAI posting $295,000 to $445,000 Master's or PhD, or senior engineers
ML infrastructure (MLOps) Run training and serving systems Anthropic posting $350,000 to $500,000 Backend and platform engineers
AI product manager Decide what the model product does Scale AI posting $171,200 to $214,000 Experienced PMs
AI safety and policy Shape AI rules and how firms meet them Scale AI posting $164,800 to $206,000 Policy, legal, government

For scale: BLS puts the median wage for all US workers at $50,980 in May 2025, and average projected growth for all occupations at 3%.

Is AI a good career in 2026?

Yes, on the government's own numbers, and more so on the data side than the research side. BLS projects data scientist employment to grow 35% from 2025 to 2035, "much faster than average", with about 24,800 openings a year. Computer and information research scientists are projected to grow 22%, but that is a field of only 38,600 jobs with about 2,900 openings a year. Software developers grow more slowly, at 10%, but from a base of about 1.7 million jobs, so the absolute gain (174,700) is the largest of the three.

BLS ties the demand directly to AI. Its research scientist profile says their expertise "will be needed in the creation of new technologies related to artificial intelligence (AI)"; the data scientist profile says firms "are expected to continue to integrate artificial intelligence (AI)-based systems into their workflows"; the software developer profile cites "the continued expansion of software development for artificial intelligence (AI)".

Our take: the growth figures are real, but so is the competition for the few roles at frontier labs. The bulk of AI hiring is ordinary companies wiring models into their products. That is where most people should aim first.

Applied AI engineer: the widest door in

An applied AI engineer (also called AI engineer, LLM engineer or forward-deployed engineer) builds products and workflows on top of existing models: retrieval, agents, evaluations, integrations. It is software development with a model in the loop, so the closest BLS line is software developers: $135,980 median in May 2025, typically a bachelor's degree.

One live example: Anthropic's "Applied AI Engineer, Enterprise" posting in San Francisco and New York lists a base salary of $200,000 to $320,000 and asks for 4+ years in roles such as forward-deployed engineer, product engineer or software engineer. That is one company's range for an experienced hire, not a starting salary.

Our take: this is the best AI job for anyone who already writes code. The skills transfer from the job you have, and every company adopting AI needs them. Our guide to the best AI for coding covers the tools these engineers use daily.

Data scientist: the fastest-growing occupation

Data scientists, in BLS's words, "use machine learning to classify or categorize data or to make predictions." Median pay was $120,230 in May 2025, about 275,600 people held the job in 2025, and the typical entry requirement is a bachelor's degree. The 35% projected growth is the highest of the three BLS profiles we read.

For a live example, Databricks' "Senior Data Scientist" posting in Mountain View and San Francisco lists a pay range of $123,200 to $218,000. That is a senior role at a data platform company, and it is a useful reality check against frontier-lab headlines.

Our take: the best entry point for people with statistics, economics or analytics backgrounds who do not want to become full-time software engineers.

Highest paying AI jobs: research scientist and research engineer

Computer and information research scientists have the highest median pay among the BLS occupations we read: $140,300 a year in May 2025 ($67.45 an hour). They typically need at least a master's degree; BLS adds that "some employers prefer to hire candidates who have a Ph.D." and that a bachelor's may be enough for some federal government jobs.

Frontier labs pay far above the median for senior people. OpenAI's "Research Engineer/Research Scientist, RL/Reasoning" posting in San Francisco lists a salary of $295,000 to $445,000 plus equity. It names no degree requirement; it asks for "a deep understanding of machine learning" and comfort "diving into a large ML codebase".

ML infrastructure and MLOps: the engineers who keep models running

Training and serving large models is a systems problem: clusters, schedulers, data pipelines, monitoring. Platform and backend engineers move into this work with the least retraining. We could not confirm a separate BLS figure for MLOps; it falls within software development. Anthropic's "Machine Learning Infrastructure Engineer, Safeguards Research" posting lists $350,000 to $500,000 base.

Our take: underrated. Fewer people chase it than research, and the skills (Linux, distributed systems, GPUs) are hard to fake.

AI product manager, safety and policy roles

Non-engineers have real routes in. Scale AI's "AI Product Manager (Coding/Multimodal)" posting lists a base salary of $171,200 to $214,000, and its "AI Policy Lead" posting lists $164,800 to $206,000; both ranges are for San Francisco, New York or Seattle. Scale says each range may span several career levels, and the policy role asks for experience with foreign governments or the diplomatic community, so neither is an entry job. We could not confirm a BLS growth figure specific to AI product or policy work.

How to get a job in AI: your first role

  1. Pick one target role from the table. Applied AI engineering and data science have the most openings; research has the fewest.
  2. Learn the core stack. Python, SQL, basic statistics, and for engineers, how to call model APIs, build retrieval and write evaluations. Our skills guide for beginners lists what to learn first.
  3. Ship two or three public projects. A deployed app with an evaluation set beats a certificate. Write down what failed and how you measured it.
  4. Apply before you feel ready. Anthropic's applied AI posting says: "We encourage you to apply even if you do not believe you meet every single qualification."
  5. Move sideways inside your company. The easiest AI job to get is the AI project at the employer you already work for.

AI jobs without a degree

Possible, but harder. BLS lists a bachelor's degree as the typical entry level for data scientists and software developers. Anthropic's applied AI posting sets the minimum as "Bachelor's degree or an equivalent combination of education, training, and/or experience", and Scale AI's product manager posting accepts "equivalent practical experience", so a strong portfolio and work history can substitute. If you want a formal route, see courses and degrees in AI.

Limits of this guide

BLS figures are US medians for broad occupations, not AI-only roles, and are projections, not guarantees. Posting ranges are six US roles at four tech firms, mostly in San Francisco, and exclude equity; most employers pay less. We did not survey pay outside the US.

Sources & Further Reading

Claim Source Date
Research scientists: $140,300 median (May 2025), master's typical, 38,600 jobs, 22% growth 2025 to 2035, 2,900 openings a year; all-workers median $50,980; average growth 3% BLS Occupational Outlook Handbook: Computer and Information Research Scientists Modified 2026-08-27; read 2026-11-17
Data scientists: $120,230 median, bachelor's typical, 275,600 jobs, 35% growth, 24,800 openings a year BLS Occupational Outlook Handbook: Data Scientists Modified 2026-08-27; read 2026-11-17
Software developers: $135,980 median, about 1.7 million jobs, 10% growth, 174,700 added BLS Occupational Outlook Handbook: Software Developers Modified 2026-08-27; read 2026-11-17
Applied AI Engineer, Enterprise (San Francisco, New York): $200,000 to $320,000 base, 4+ years; degree "or equivalent" wording; "apply even if" wording Anthropic posting 5057647008 Read 2026-11-17
Machine Learning Infrastructure Engineer, Safeguards Research (San Francisco, New York): $350,000 to $500,000 base Anthropic posting 5364804008 Read 2026-11-17
Research Engineer/Research Scientist, RL/Reasoning (San Francisco): $295,000 to $445,000 salary, offers equity; requirements quoted OpenAI posting 7e198d81 Read 2026-11-17
Senior Data Scientist (Mountain View, San Francisco): $123,200 to $218,000 local pay range Databricks posting 5634684002 Read 2026-11-17
AI Product Manager (Coding/Multimodal): $171,200 to $214,000 base (San Francisco, New York, Seattle); "equivalent practical experience" wording; range may span career levels Scale AI posting 4609736005 Read 2026-11-17
AI Policy Lead: $164,800 to $206,000 base (San Francisco, New York, Seattle); foreign-government experience Scale AI posting 4752554005 Read 2026-11-17

Frequently asked questions

What are the best AI jobs in 2026, and is AI a good career?

The best AI jobs for most people in 2026 are applied AI engineer and data scientist. BLS projects data scientist jobs to grow 35% from 2025 to 2035 at a $120,230 median; research scientists earn more ($140,300) but usually need a master's.

What is the highest paying AI job?

In BLS data, computer and information research scientists have the highest median pay of the AI-related occupations we read: $140,300 in May 2025. Individual postings at frontier labs go far higher; one OpenAI research engineer and scientist posting in San Francisco lists $295,000 to $445,000 plus equity.

Can you get an AI job without a degree?

Sometimes. BLS says data scientists and software developers typically need a bachelor's degree. Anthropic's applied AI posting accepts "a Bachelor's degree or an equivalent combination of education, training, and/or experience", so shipped work and years in technical roles can stand in for the diploma.

Do you need a PhD to work in AI research?

Not always. BLS says computer and information research scientists typically need at least a master's degree, and some employers prefer a Ph.D. A bachelor's may be enough for some federal government jobs. OpenAI's RL research posting we read lists no degree; it asks for a deep understanding of machine learning.

How we sourced this. We read three US Bureau of Labor Statistics Occupational Outlook Handbook profiles (modified 2026-08-27) and six job postings from Anthropic, OpenAI, Databricks and Scale AI on 2026-11-17. Methodology