We scanned 195,955 US job postings for AI skill requirements. 16.2% ask for AI, yet only 3.5% are AI-native roles. Most of the demand is AI spreading into ordinary jobs, and it is heavily concentrated by field.
The Headline Numbers
Methodology
We scanned 195,955 active US job postings and flagged each one that requires AI: either the role is AI-native, like a machine learning engineer, or the description asks for AI tools and skills, like a marketing manager expected to use generative AI. We split those two so the AI-native share does not get inflated by ordinary jobs that merely mention AI.
Category adoption is the share of postings in that field requiring AI. Fields with fewer than 500 postings are excluded. The figures refresh on every build and reflect the week dated 2026-06-08.
The Data
| Field | Require AI | AI postings | Total postings |
|---|---|---|---|
| Engineering | 45.6% | 10,639 | 23,343 |
| Product | 37.6% | 2,187 | 5,821 |
| Data | 26.6% | 3,072 | 11,556 |
| Marketing | 21.8% | 1,758 | 8,072 |
| Finance | 13.1% | 929 | 7,105 |
| People | 12.0% | 505 | 4,208 |
| Sales | 11.3% | 4,135 | 36,505 |
| Legal | 10.9% | 321 | 2,953 |
| Operations | 9.5% | 1,084 | 11,433 |
| Other | 8.3% | 7,025 | 84,959 |
A posting counts as requiring AI when the role is AI-native or the description asks for AI tools or skills. From 195,955 2026 US postings. Fields with fewer than 500 postings are excluded.
What the numbers say
The number that matters is the gap between 16.2% and 3.5%. Roughly one in six postings asks for AI skills, but only 3.5% are dedicated AI roles. The other AI demand is sitting inside ordinary job descriptions: the marketer expected to run generative campaigns, the analyst expected to prompt their way through a dataset. AI has become a line item in the requirements rather than a separate career.
It lands unevenly. Engineering postings ask for AI 5.5 times as often as other postings. That gap is the practical map of where AI fluency has already become table stakes and where it is still an edge. If your field is near the top, the skill is becoming a baseline you cannot skip. If it is near the bottom, learning it now is how you get ahead of your peers before it becomes the baseline there too.
FAQ
16.2% of 195,955 US job postings we analyzed require AI skills, meaning the role is AI-native or the description asks for AI tools. Only 3.5% are AI-native roles, so most of the demand is AI spreading into existing jobs.
Engineering leads at 45.6% of postings, then Product at 37.6% and Data at 26.6%. Other is lowest at 8.3%.
Mostly changing existing ones. Only 3.5% of postings are AI-native roles, while 16.2% require AI in total. The difference is ordinary jobs that now expect AI skills in the description.
It is the share of active US job postings that require AI, counted across 195,955 2026 postings. A posting counts when the role is AI-native or the description asks for AI tools or skills. Fields with fewer than 500 postings are excluded.
AI Pulse