Half of US content design and UX writing jobs ask for AI tools. In technical writing, 1 in 10 do
We read 8,709 US content job postings first seen on company career sites Oct 1 to Nov 4, 2026. 20% ask the worker to use AI tools. The ask is strongest in content design and UX writing (50%) and SEO (48%), and weakest in technical writing (9%).
What we found
- Content design and UX writing leads. 50% of its 113 postings ask for AI tools, against 20% across content jobs and 9% for technical writing (714 postings).
- SEO now means writing for AI answers too. 45% of SEO postings ask for content that shows up in AI search (AEO, GEO, AI Overviews), and 56% ask for AI tools or AI search. Across all content jobs, 5% mention AI search.
- The roles that ask most for AI also pay more. Content design and UX writing postings list the highest median posted pay, $140K, and 50% ask for AI tools. Video and multimedia lists the lowest, $75K, with 19% asking.
- Inside a role, postings that ask for AI tools list more pay than postings that never name AI in 8 of the 8 roles with enough posted pay to compare (median gap +30%). These are different employers, so the gap may say more about who asks than about the ask.
- Most postings say "AI tools" without naming one. 4.7% name a tool; ChatGPT is named most (284 postings), then Claude, Gemini and Perplexity.
- AI training gigs are counted apart. 149 postings pay writers and experts to write questions for AI models or grade their answers, 90% of them from one AI training platform. 96% are remote and the median posted rate is $69 an hour. Mixed in, they would inflate both the AI share and the remote share of content work.
How we made it
Company career sites: US postings first seen Oct 1 to Nov 4, 2026. Content jobs are found by job title in ten groups (content design and UX writing, SEO, content strategy, copywriting, content marketing, video and multimedia, social media, editing, writing, technical writing); internships, student jobs, titles naming a non-US place, staffing agencies and job aggregators are left out.
One posting = one employer + job title + city; copies on several career sites of one employer count once (9,524 rows became 9,449 postings). 8,709 content postings whose text we could read are charted. No employer has more than 13% of a role's postings; keeping at most 10 postings per employer gives 21% asking for AI tools (all postings: 20%).
Asks to use AI tools: a sentence asks the worker to use AI (AI tools, AI-assisted drafts, AI image or video tools, ChatGPT, Claude and the like, prompting, or experience with AI), or the title names AI. Writes for AI search: AEO, GEO, AI Overviews, visibility in AI answers. Sentences about AI in the hiring process (notices, privacy, screening) do not count; 1.1% of postings name AI only there. 63 postings from 11 employers forbid AI in the work; they count as naming AI, not as asking for it.
Pay: stated pay only (no estimates), USD, the midpoint of the range, made yearly (hourly x 2080, daily x 260, weekly x 52, monthly x 12), kept between $20K and $1M; a median is shown only where at least 30 postings state pay. Remote: fully remote, read from the posting's place and work type the way the rest of the site does.
AI training gigs (question writers, content evaluators, AI trainers) are a separate group and not in the content totals. These are postings, not hires, and the text rules are keyword rules: they can miss an ask worded another way.
Run your own study
This research uses the same live job postings you can search through Treg: open roles read daily from employers' own career sites.
Related research


