OpenAI NewsJul 27, 2026, 11:30 UTC+8

OpenAI study: 43.5% of occupation-specific ChatGPT use crosses job boundaries

How AI is expanding what people do at work

OpenAI Economic Research analyzed 800,000+ ChatGPT messages from US users. 16.8% of work messages and 43.5% of occupation-specific messages involve tasks from another occupation—a pattern they call 'task crossover.' Customer experience (77%), design (75%), and HR (69%) workers borrow the most. Marketing and engineering tasks travel farthest across fields. Crossover is more common in small businesses. The report also notes AI is creating new tasks like prompt engineering and output review that don't fit standard job classifications. This is the first paper in the 'Work at the Frontier' series; the full PDF is available.

Why it matters: OpenAI's own research with 800k conversations as the dataset—credible scale. The 43.5% crossover rate is a fresh signal, far more concrete than generic 'AI changes work' narratives. Not an 85 because it's a report, not a product launch or model release—impact is more diffuse.

Read the original ↗Export Markdown

Take

This one's worth opening because it shifts the AI-and-jobs conversation from "which roles disappear" to "tasks are leaking across job boundaries." OpenAI's economic research team took 800K+ ChatGPT messages from US users, stripped out generic stuff like writing emails and summarizing, and found that 43.5% of the remaining occupation-specific messages involved tasks from a different occupation. Customer experience workers (77%), designers (75%), and HR (69%) borrow the most. Marketing and engineering tasks travel farthest—they show up in conversations from people outside those fields. Crossover is more common in small businesses, which makes sense: fewer people means everyone wears more hats.

I'd discount the number a bit. The data is ChatGPT users only, not the whole workforce, and the post doesn't detail how they classified which task belongs to which occupation—that boundary-drawing directly affects the 43.5% figure. But the direction is real: AI lets non-coders tweak websites and non-writers produce copy, so job boundaries are softening. The report also flags that new tasks like prompt engineering and output review don't fit existing job classifications, which is a signal worth tracking.

Source: OpenAI News · openai.com
Published: Jul 27, 2026, 11:30 UTC+8 · Jul 27, 2026