Studies how to make advanced AI systems behave safely and as intended, through research on alignment, interpretability and dangerous capabilities. BLS does not publish separate figures for computer and information research scientists; it counts them within computer and information research scientist (SOC 15-1221), whose median pay is $140,300 a year (BLS OEWS, May 2025). Typical entry-level education: Master's degree.
- SOC 15-1221
- Computer and Mathematical
- Last updated October 9, 2026
At a glance
AI Safety Researcher in four numbers
- Median pay
- $140,300a year, or $67.45/hrBLS OEWS, May 2025Lowest 10% $82KTop 10% $231K
US median, all jobs: $50,980
- Projected growth
- +21.8%Much faster than averageBLS EP, 2025–2035This career+21.8%All jobs+3.5%
Projected employment change, 2025–2035 Item Value This career +21.8% All jobs +3.5% - Openings a year
- 2,90038,600 jobs todayBLS EP, 2025–2035
- Jobs in 202538.6K
- Projected 203547K
Employment, 2025–2035 Item Value Jobs in 2025 38.6K Projected 2035 47K - Typical education
- Master'sneeded to enterBLS EP, 2025–2035; mix: O*NET
- Some college8.4%
- Bachelor's36.2%
- Master's19.9%
- Professional degree2.6%
- Doctorate28.4%
- Post-doctoral training4.5%
Highest education held by computer and information research scientists Segment Value Some college 8.4% Bachelor's 36.2% Master's 19.9% Professional degree 2.6% Doctorate 28.4% Post-doctoral training 4.5%
Also advertised as: Alignment Researcher, AI Alignment Research Scientist, Interpretability Researcher, AI Safety Research Engineer.
Pay
How much do Computer and Information Research Scientists make?
Every ai safety researcher in the US, by percentile of annual wage
Median
$140,300
Middle half earn $103,570 to $188,700
- 10th
- $82,200
- 25th
- $103,570
- 75th
- $188,700
- 90th
- $230,630
All US occupations: $50,980
| Percentile | Value |
|---|---|
| 10th | $82,200 |
| 25th | $103,570 |
| Median | $140,300 |
| 75th | $188,700 |
| 90th | $230,630 |
| All US occupations | $50,980 |
Source: BLS OEWS, May 2025
- Median annual wage
- $140,300
- Median hourly wage
- $67.45/hr
- Mean annual wage
- $153,930
- People employed
- 37,200
Versus all US occupations
At $140,300, Computer and Information Research Scientists earn 175% above the $50,980 median across all US occupations, a difference of $89,320 a year.
Versus computer and mathematical occupations
Inside the Computer and Mathematical group, where the middle occupation pays $116,580, this is 20% above the group median.
Pay range
The top 10% earn $230,630 or more while the bottom 10% earn $82,200, a gap of 2.8×. That is a typical spread for a US occupation.
Rank in the bank
This median pays more than 91% of the 1,741 occupations in the NueCareer career bank.
Highest-paying state
Oregon pays the most at $206,220, and still leads at $199,514 once local prices are taken into account.
Pay by state
Which states pay Computer and Information Research Scientists the most?
State medians come from the same OEWS release as the national figures. The cost-of-living column divides each state median by that state's BEA Regional Price Parity, restating pay in national-average dollars so a high-cost state and a low-cost one can be compared directly.
| State | Median | COL-adjusted | Employed |
|---|---|---|---|
| 1Oregon | $206,220 | $199,514 | 730 |
| 2Washington | $202,400 | $189,136 | 2,050 |
| 3New York | $181,990 | $168,633 | 1,950 |
| 4Massachusetts | $170,510 | $161,228 | 1,760 |
| 5New Mexico | $165,690 | $179,684 | 490 |
| 6California | $159,100 | $143,696 | 8,440 |
| 7Virginia | $157,390 | $155,671 | 3,590 |
| 8Maryland | $145,310 | $138,445 | 2,870 |
| 9District of Columbia | $144,190 | $131,200 | 330 |
| 10New Jersey | $140,840 | $129,443 | 1,580 |
Job outlook
Is AI Safety Researcher a growing career?
Employment change over 2025–2035, against every US occupation
Much faster than average. +21.8% projected growth vs 3.5% for all occupations.
| Item | Value |
|---|---|
| Computer and Information Research Scientists | +21.8% |
| All US occupations | +3.5% |
Source: BLS EP, 2025–2035
Much faster than average
BLS projects employment of Computer and Information Research Scientists to change 21.8% over 2025–2035, roughly 6.2× the 3.5% projected for all occupations. That is 8,400 more jobs.
Annual openings
About 2,900 openings are projected each year, 7.5% of the 38,600 jobs that existed in 2025. Most come from workers leaving the occupation, not from growth alone.
Getting in
How to become an AI Safety Researcher
Many AI safety researchers hold a doctorate in computer science, mathematics, physics or a related field, but a large share of the field enters with a bachelor's or master's degree and strong machine learning engineering skills. Practical experience training and probing neural networks matters more than a specific degree.
Common routes include research fellowships and residency programs run by AI labs and safety organizations, contributing to published safety research, and replicating interpretability or evaluation papers. Government bodies such as the NIST Center for AI Standards and Innovation also hire technical staff. Writing up independent research publicly is a recognized way to get noticed.
- 1
Typical entry-level education
Master's degree
- 2
Work experience in a related occupation
None
- 3
Typical on-the-job training
None
- 4
Preparation needed (O*NET job zone)
Considerable preparation
BLS Employment Projections, 2025–2035; job zone from O*NET 31.0 Database.
What education people in the job have
O*NET incumbent survey: the highest level of education reported by people already working as computer and information research scientists.
- Bachelor's Degree36.2%
- Doctoral Degree28.4%
- Master's Degree19.9%
- Some College Courses8.4%
- Post-Doctoral Training4.5%
| Item | Value |
|---|---|
| Bachelor's Degree | 36.2% |
| Doctoral Degree | 28.4% |
| Master's Degree | 19.9% |
| Some College Courses | 8.4% |
| Post-Doctoral Training | 4.5% |
Source: O*NET 31.0 Database
The work
What does an AI Safety Researcher do?
AI safety researchers work on making powerful AI systems reliable, honest and controllable. Some focus on alignment, developing training methods that keep models following human intentions. Others work on interpretability, reverse engineering what is happening inside a neural network, or on evaluations that test whether a model has dangerous capabilities, such as helping with cyberattacks or deceiving its users.
The work combines research and engineering. Researchers form hypotheses, run experiments on real models, analyze results and publish papers or internal reports. At AI labs, their findings shape decisions about whether and how to release a model. At government institutes and nonprofits, they test models from many companies and inform policy.
Day-to-day tasks
- Design experiments that test how models behave in risky situations
- Develop training methods that make models more honest and controllable
- Analyze the internal workings of neural networks
- Build evaluations for dangerous capabilities such as cyber misuse
- Write papers and internal reports on research findings
- Advise model release decisions with evidence from testing
- Review and replicate other researchers' safety results
Typical tasks, researched by NueCareer from the sources below.
Working conditions and hours
AI safety researchers work at AI labs, university groups, nonprofits and government institutes, concentrated in the San Francisco Bay Area, London and Washington. The work is desk-based, with flexible schedules, frequent reading groups and collaborative research. Some roles involve testing models under tight pre-release deadlines.
Where the work happens
Indoors, climate-controlled
NeverEvery dayIndoors, in a climate-controlled space, every day.
Outdoors in all weather
NeverEvery dayAlmost never works outdoors.
Physical demands
Sitting
NeverAlmost all the timeSeated for more than half the working day.
Standing
NeverAlmost all the timeBending or twisting
NeverAlmost all the time
Working with people
Contact with other people
NoneConstantFace-to-face discussions
NeverEvery dayFace-to-face discussions every day.
Dealing with the public
Not importantExtremely important
Pressure and stakes
Strict deadlines
NeverEvery dayConflict situations
NeverEvery dayCost of a mistake
Not seriousExtremely serious
Autonomy
Freedom to decide
NoneA lotA lot of freedom to make decisions without checking first.
Conditions: O*NET Work Context survey. Each item has its own scale, so each dot is read against the two answers that end its own line.
Core skills
O*NET importance, 1 to 5
- Critical Thinking4.0
- Reading Comprehension3.8
- Active Listening3.8
- Speaking3.5
- Active Learning3.5
- Writing3.4
- Mathematics3.4
- Science3.1
| Item | Value |
|---|---|
| Critical Thinking | 4.0 |
| Reading Comprehension | 3.8 |
| Active Listening | 3.8 |
| Speaking | 3.5 |
| Active Learning | 3.5 |
| Writing | 3.4 |
| Mathematics | 3.4 |
| Science | 3.1 |
Source: O*NET 31.0 Database
Transferable skills
O*NET importance, 1 to 5
- Complex Problem Solving4.0
- Judgment and Decision Making4.0
- Systems Analysis3.8
- Programming3.6
- Systems Evaluation3.6
- Time Management3.4
- Operations Analysis3.1
- Technology Design3.1
| Item | Value |
|---|---|
| Complex Problem Solving | 4.0 |
| Judgment and Decision Making | 4.0 |
| Systems Analysis | 3.8 |
| Programming | 3.6 |
| Systems Evaluation | 3.6 |
| Time Management | 3.4 |
| Operations Analysis | 3.1 |
| Technology Design | 3.1 |
Source: O*NET 31.0 Database
Tools and software
- Analytical or scientific software · in demand
- Application server software · in demand
- Business intelligence and data analysis software · in demand
- Cloud-based management software · in demand
- Configuration management software · in demand
- Data base management system software · in demand
- Data base user interface and query software · in demand
- Data mining software · in demand
- Development environment software · in demand
- Enterprise application integration software · in demand
- Expert system software · in demand
- File versioning software · in demand
“In demand” marks an O*NET Employer-Based Hot Technology, software named in real job postings for this occupation.
Personality fit
What personality type suits an AI Safety Researcher?
NueCareer estimate (Holland code IRS) built on O*NET's Computer and Information Research Scientist profile, scored 1 to 7
| Axis | AI Safety Researcher |
|---|---|
| Realistic | 5 |
| Investigative | 7 |
| Artistic | 3 |
| Social | 3 |
| Enterprising | 2 |
| Conventional | 5 |
Source: O*NET 31.0 Database
What the work is built around
Day to day this is work built around investigating, analysing and solving problems with ideas and data, combined with organising, following procedures and working precisely with data.
- Investigative7.0
- Conventional4.8
- Realistic4.5
- Social3.1
- Artistic3.0
- Enterprising2.5
Closest MBTI types
NueCareer fit scores: cosine similarity between each MBTI type's interest vector and this occupation's O*NET interest profile. Not a BLS or O*NET figure; the method is documented on our methodology page.
Questions
AI Safety Researcher: frequently asked questions
How much do Computer and Information Research Scientists make?
BLS does not publish a separate wage for computer and information research scientists. It counts them within computer and information research scientist, whose median pay is $140,300 a year (BLS OEWS, May 2025). The middle half of that occupation earn between $103,570 and $188,700, and the top 10% earn $230,630 or more.
Do you need a degree to become an AI Safety Researcher?
Yes. BLS reports a master's degree as the typical entry-level education for Computer and Information Research Scientists, and O*NET rates the preparation needed as considerable preparation. In O*NET's survey of people already doing the job, 36.2% hold bachelor's degree.
Is AI Safety Researcher a growing career?
Yes, strongly. Employment is projected to grow 21.8% over 2025–2035, more than double the 3.5% projected for all occupations (BLS EP, 2025–2035). BLS still projects about 2,900 openings a year, mostly from workers retiring or moving to other work.
What does an AI Safety Researcher do?
AI safety researchers work on making powerful AI systems reliable, honest and controllable. A typical task: design experiments that test how models behave in risky situations. Employers most often ask for Analytical or scientific software, Application server software, Business intelligence and data analysis software.
Which states pay Computer and Information Research Scientists the most?
The highest state medians for Computer and Information Research Scientists are Oregon ($206,220), Washington ($202,400), New York ($181,990) (BLS OEWS, May 2025).
What personality type suits an AI Safety Researcher?
O*NET scores Computer and Information Research Scientists highest on the Investigative, Conventional and Realistic interest areas, so the work mixes investigating, analysing and solving problems with ideas and data with organising, following procedures and working precisely with data. On NueCareer's MBTI-to-interest mapping that lines up most closely with INTJ personalities. Personality is a fit signal, not a gate. The free NueCareer quiz scores your own profile against all 1,731 careers in the bank.






