Do You Need a Degree to Work in AI? What the Data Shows

What the federal data says, role by role

BLS does not track AI jobs as a separate occupation, so you have to read the nearest series. Machine learning engineers, AI developers, agent developers, RPA developers and most other AI building roles fall under software developers. For that group, BLS gives a bachelor's degree as the typical entry-level education, usually in computer and information technology, engineering or mathematics. It adds that some employers prefer a master's degree.

Data scientists also have a bachelor's degree as the typical entry level, though BLS says some jobs require a master's or doctoral degree. Research is the exception. For computer and information research scientists, the series closest to AI research scientists, BLS lists a master's degree as typical and says some employers prefer a Ph.D. It also notes that the federal government may hire some research candidates with a bachelor's degree.

O*NET puts software developers, data scientists and computer systems analysts in Job Zone Four. Its description of that zone is that most of these jobs require a four-year bachelor's degree, but some do not. Computer and information research scientists sit in Job Zone Five, where most jobs require graduate school.

How many people in these jobs have no degree

The "typical entry-level education" field shows what employers usually expect. It does not tell you who actually holds the jobs. BLS also publishes the education levels of workers aged 25 and over in each occupation, based on 2023-24 data. In that table, 14.0% of software developers have no bachelor's degree, and 34.2% have a master's or doctorate.

Some of the other series sit further from the degree norm. Among computer systems analysts, the nearest series to workflow automation, 27.9% have no bachelor's degree. The figure is 26.1% for computer and information systems managers, 21.7% for database architects and 21.0% for management analysts, the series this site uses for AI consultants. In computer occupations, all other, the closest series to solutions architects, 43.4% have no bachelor's. Research is the hardest to enter without a degree: only 6.2% of computer and information research scientists lack a bachelor's, and 54.7% hold a graduate degree.

Where the evidence is thin

Every figure on this page comes from a broad BLS or O*NET occupation that covers the whole field. None of them looks at AI specialists alone. Many of the workers without degrees may have been hired years ago, into roles where the bar was lower. We found no federal data on how often employers hire people without degrees into AI-specific roles today, so treat those percentages as a ceiling on what is possible. They are not a forecast of your chances.

Worker education data also can't separate what an employer requires from what candidates happen to bring. A posting that lists a degree as preferred may still hire someone without one who can show the work. In research roles, the high share of graduate degrees reflects how that work is actually done, and few teams make exceptions.

What to do if you don't have a degree

Aim for the roles where the data shows the most flexibility: applied engineering, automation and integration work, and systems analysis. Proof of skill matters most here. That means code you can show, systems you have shipped, and the ability to explain your tradeoffs in a technical interview. Leave research roles until later, or plan to earn a graduate degree if research is your goal.

Registered apprenticeships are one structured route in. O*NET lists Data Scientist and Machine Learning Data Curator as titles approved by the U.S. Department of Labor for Registered Apprenticeship Programs. Certifications can show you know a specific platform. They do not stand in for a degree in a screening filter, and no certification guarantees a job.

What you need

No licenseNo license or regulator governs AI work itself. Hiring depends on what each employer requires.
Bachelor's, typical for engineeringBLS lists a bachelor's degree as the typical entry-level education for software developers, the nearest series to most AI engineering roles. Some employers prefer a master's.
Bachelor's, typical for data scienceBLS lists a bachelor's degree for data scientists, usually in math, statistics, computer science, business or engineering. Some jobs require a master's or doctorate.
Master's, typical for researchBLS lists a master's degree for computer and information research scientists, and says some employers prefer a Ph.D.
Shown programming skillBLS says software developers need a strong programming background and must keep learning new tools and languages throughout their careers.

How to get there

Month 1Pick a target roleApplied engineering, automation and analyst roles show the most workers without degrees in BLS data. Research shows the fewest.
Months 1-12Build core skillsLearn programming, data handling and the math your target role uses, whether through a degree, self-study or a registered apprenticeship.
Months 6-18Ship public workBuild and document projects that solve real problems, so an employer can judge your skill without leaning on your credentials.
Year 1-2Get a first technical jobMany people start in an adjacent software, data or automation job and move into AI work from inside the company.
Year 2+Revisit graduate studyIf you want research roles, plan for a master's or Ph.D. BLS lists a master's degree as typical for that work.

Open now on AutomationRoles.ai

Manager, Data Scientist - Business Cards & PaymentsCapital One · New York, New York

Principal Associate, Data Scientist - SBB Marketing & Deposit StrategyCapital One · McLean, Virginia

Manager, Data Scientist - OptimizationCapital One · New York, New York

Senior Data Scientist, AI FoundationsCapital One · New York, New York

Manager, Data Scientist - SBB Marketing & Deposit StrategyCapital One · McLean, Virginia

AI Engineer (AI Accelerator Program)UCSF · San Francisco, California

See all →

Questions people ask

Can you become a machine learning engineer without a degree?

It is possible, but it means going against the usual pattern. BLS lists a bachelor's degree as typical for software developers, the nearest series. Even so, 14% of workers aged 25 and over in that occupation have no bachelor's degree.

Do AI research jobs need a PhD?

BLS lists a master's degree as the typical entry-level education for computer and information research scientists, and says some employers prefer a Ph.D. In 2023-24, 23.8% of workers aged 25 and over in that occupation held a doctoral or professional degree, and 30.9% held a master's.

Do data scientists need a master's degree?

Not usually at entry level. BLS gives a bachelor's degree as typical for data scientists, but it notes that some jobs require a master's or doctoral degree.

Are there apprenticeships in AI or data science?

Yes. O*NET lists Data Scientist and Machine Learning Data Curator as titles approved by the U.S. Department of Labor for Registered Apprenticeship Programs. Openings vary by employer and location.

Can a certification replace a degree?

No federal data shows that a certification substitutes for a degree in hiring. A certification can show you know a specific platform, but it won't get you hired by itself.

Do you need a degree to work in AI?