Study · AI pay premium · 3 November 2026
Do AI jobs pay more?
US job ads with AI or machine learning in the title advertise a median of $230,825 a year, 26% more than other ads. But about half of that gap is who is hiring and at what level. At the same employer and level, AI roles post 13.2% more than the employer's other roles, and AI engineers are offered just 1.4% more than its software engineers. From 36,963 US job ads on the public job boards of 529 employers, read on 3 November 2026.
By Dominic Roworth · Data as of 3 November 2026, refreshed every month · Download the data (CSV)
AI in the title vs all other ads
+26%
$230,825 vs $183,750, median posted midpoint
Same employer, same level
+13.2%
AI roles vs the employer's other roles, 426 comparisons
AI engineers vs software engineers
+1.4%
Same employer and level, 230 comparisons
The headline gap
1,823 of the 36,963 US job ads in the snapshot (4.9%) had AI or machine learning in the title, at 307 of the 529 employers. Counting each distinct role once, 1,426 of them published a yearly base pay range. The typical midpoint of those ranges was $230,825, against $183,750 for the 21,560 other roles with a yearly range: 26% more.
That is the number that makes headlines. It mixes up three things: the job, the level, and the employer.
Part of it is who is hiring
The employers that advertised AI roles pay more for everything. Their ads for jobs without AI in the title had a median midpoint of $185,900, against $169,500 at employers with no AI ads: 10% more before a single AI job is counted. AI roles are also a little more senior: of the ads whose title names a level, 93.1% of AI ads were senior, lead or manager roles, against 88.5% of other ads.
Same employer, same level
So we compared each employer's AI roles with its own other roles at the same level (senior with senior, and so on), wherever it advertised both: 426 comparisons. The typical AI role posted 13.2% more. AI paid more in 313 of them and less in 74. That is about 48% less than the headline gap: roughly half of the "AI premium" in job ads is who is hiring and at what level.
AI engineers vs software engineers
The closest comparison is between engineers. Where one employer advertised both AI engineering roles (AI, machine learning or ML engineers and developers) and software engineering roles at the same level, the typical AI role's midpoint was 1.4% higher, across 230 comparisons. The AI role paid more in 126, less in 63 and the same in 41.
For an engineer, then, the money in AI is mostly in getting hired by the employers that pay the most, not in the job title. Inside a company, AI engineers and software engineers are usually on the same pay scale.
The AI premium shrinks as the comparison gets closer
How much higher the typical AI role's posted midpoint is, against other roles.
- AI in the title vs all other ads +26%
- Same employer, same level +13.2%
- AI engineers vs software engineers, same employer and level +1.4%
By level
Median midpoint of yearly ranges, all employers, by the level the title names. Levels with fewer than 30 AI roles are left out. These still mix employers, so the gaps are wider than within one company.
| Level | AI roles | AI median | Other roles | Other median | Gap |
|---|---|---|---|---|---|
| Senior, staff or principal | 714 | $234,750 | 7,912 | $193,050 | +21.6% |
| Lead | 63 | $237,500 | 1,079 | $195,900 | +21.2% |
| Manager, director or above | 131 | $272,500 | 3,143 | $221,875 | +22.8% |
| Title names no level | 489 | $216,104 | 8,796 | $160,000 | +35.1% |
The most common AI job titles
Titles with level words removed (so a Senior Machine Learning Engineer counts as a machine learning engineer), with at least 15 distinct roles. "Software engineer" here means a software engineering role with AI in the title, such as "Software Engineer, Applied AI".
- Member of technical staff $293,750
- Applied ai architect $284,500
- Engineering manager $274,399
- Applied ai engineer $249,750
- Machine learning engineer $247,500
- Ml engineer $240,000
- Product manager $228,500
- Ai engineer $225,000
- Software engineer $215,500
| Title | Roles with a yearly range | Median posted midpoint |
|---|---|---|
| Software engineer | 224 | $215,500 |
| Machine learning engineer | 199 | $247,500 |
| Product manager | 64 | $228,500 |
| Ai engineer | 51 | $225,000 |
| Applied ai engineer | 32 | $249,750 |
| Engineering manager | 20 | $274,399 |
| Member of technical staff | 18 | $293,750 |
| Ml engineer | 16 | $240,000 |
| Applied ai architect | 15 | $284,500 |
Where the AI jobs are, by sector
Sectors with at least 30 AI ads. Share is the share of the sector's US ads with AI in the title. The median is shown where at least 30 AI roles had a yearly range.
- Ecommerce tech 18.4%
- Autonomous vehicles 16.2%
- Consumer 11.2%
- AI 10.9%
- Data and analytics 9.5%
- Enterprise software 6.9%
- Hardware and chips 6.5%
- Trading 6.3%
- Cloud and infrastructure 5.8%
- Biotech 5.5%
- Fintech 4.9%
- Developer tools 4.7%
- Logistics 4.7%
- Security 4.6%
- Health tech 3.1%
- Aerospace and defence 2%
| Sector | US ads | AI ads | AI share | AI median posted midpoint |
|---|---|---|---|---|
| AI | 3,217 | 352 | 10.9% | $254,500 |
| Aerospace and defence | 9,403 | 184 | 2% | $192,750 |
| Autonomous vehicles | 910 | 147 | 16.2% | $256,000 |
| Fintech | 2,958 | 144 | 4.9% | $233,150 |
| Consumer | 1,134 | 127 | 11.2% | $253,075 |
| Data and analytics | 1,110 | 105 | 9.5% | $229,750 |
| Enterprise software | 1,306 | 90 | 6.9% | $220,000 |
| Health tech | 2,256 | 69 | 3.1% | $205,000 |
| Hardware and chips | 922 | 60 | 6.5% | Too few ranges |
| Developer tools | 1,190 | 56 | 4.7% | $221,175 |
| Security | 1,216 | 56 | 4.6% | $212,000 |
| Cloud and infrastructure | 865 | 50 | 5.8% | $215,000 |
| Logistics | 1,013 | 48 | 4.7% | $215,000 |
| Ecommerce tech | 201 | 37 | 18.4% | Too few ranges |
| Trading | 555 | 35 | 6.3% | Too few ranges |
| Biotech | 596 | 33 | 5.5% | $205,000 |
Employers with the most AI job ads
Open US ads with AI or machine learning in the title on 3 November 2026, and their share of the employer's US ads. A count of ads, not of hires.
| Employer | AI ads | Share of its US ads |
|---|---|---|
| SpaceX | 101 | 3.5% |
| Waymo | 71 | 22.8% |
| OpenAI | 63 | 10% |
| Anthropic | 54 | 10.7% |
| Databricks | 43 | 8.9% |
| Anduril | 37 | 1.5% |
| Snowflake | 32 | 13.9% |
| Samsara | 30 | 11.8% |
| Scale AI | 29 | 24.2% |
| 28 | 21.4% | |
| 26 | 21.3% | |
| Zoox | 26 | 11.7% |
| Klaviyo | 25 | 19.5% |
| DoorDash | 23 | 5.3% |
| Cerebras | 22 | 22.2% |
How we did it, and what it can't tell you
- The ads. Every open role on the public job boards (Greenhouse, Lever and Ashby) of 529 US-hiring employers, read on 3 November 2026: 50,727 ads, 36,963 of them in the US. The same snapshot runs on the first of every month, so this page will show whether the gap grows or shrinks.
- What counts as an AI job. A title with AI, artificial intelligence, machine learning, ML, deep learning, LLM or generative AI in it. A role on an AI team counts; an AI job whose title doesn't say so doesn't.
- Pay. The midpoint of a yearly base pay range in US dollars, as the employer published it, read the same way as on our job board. Ranges that include commission, hourly rates and ranges that look like typing errors are left out. Equity and bonuses aren't in posted ranges, and at AI labs they can be large. An employer that posts the same role in many places counts once.
- Not a random sample. These are mostly technology, venture-backed and fast-growing companies. Read the figures as "among employers like these", not as all US employers. For the pay of all US software developers, see the official figures on our software engineer salary page and machine learning engineer page.
- Levels come from the title (senior, staff, lead, director and so on). About half of titles name no level; they are compared with each other.
The data is free to reuse under CC BY 4.0: download the CSV, or cite the archived copy on Zenodo, DOI 10.5281/zenodo.24167606. From the same job ads: our pay transparency index, and the pay each fast-growing company posts on Hiring Radar.
Questions
Do AI jobs pay more than other jobs?
In job ads, yes. US ads with AI or machine learning in the title advertised a median of $230,825 a year (the midpoint of the posted range), against $183,750 for all other ads with a yearly range: 26% more. About 48% of that gap disappears when each AI role is compared with the same employer's other roles at the same level.
Do AI engineers earn more than software engineers?
Barely, when the employer is the same. Across 230 cases where one employer advertised both AI engineering roles and software engineering roles at the same level, the typical AI role's posted midpoint was 1.4% higher. The AI role paid more in 126 cases, less in 63 and the same in 41.
Why do AI job ads show higher salaries?
Largely because of who is hiring. The 307 employers that advertised AI roles also posted more for their other jobs: a median of $185,900 against $169,500 at employers with no AI ads, 10% more. AI roles are also slightly more senior: 93.1% of AI ads that name a level are senior, lead or manager roles, against 88.5% of other ads.
How many job ads are for AI roles?
4.9% of the 36,963 US job ads in our 3 November 2026 snapshot had AI or machine learning in the title: 1,823 ads at 307 of the 529 employers we follow.