OpenAI Unlocking New Ways of Working Report Shows Cross‑Occupation AI Use Becomes Recurring

TL;DR

OpenAI’s latest research shows that workers are repeatedly using AI for tasks outside their traditional roles, and these cross‑occupation activities are becoming a regular part of their workflows, hinting at job expansion even when job titles remain unchanged.

Cross‑Occupation AI Use Is Growing

Analyzing over 1.5 million work‑related ChatGPT messages from April to July 2026, OpenAI identified a rise in “task crossover” – workers applying AI to activities historically associated with other occupations. The share of occupation‑specific AI activity that involved cross‑occupation tasks grew from 13.1 % in April to 25.9 % in July among a cohort of roughly 6,200 consistently observed workers.

Workers Prompt AI Differently for Out‑of‑Role Tasks

When tackling tasks outside their primary domain, workers tend to:

  • Write shorter prompts than for within‑role tasks.
  • Provide more context (examples, background documents) rather than asking for explanations or step‑by‑step guidance.
  • Request verification or checks rather than seeking advice on how to perform the task.

“Workers bring a problem and relevant context—such as a document, example, or information from a colleague—and ask AI to help apply knowledge associated with another field.” – OpenAI research note

These patterns suggest that employees are using AI to borrow expertise rather than to learn a new discipline from scratch.

Recurrence Demonstrates Emerging AI‑Powered Workflows

A follow‑up analysis of one‑month intervals revealed that workers who had used a cross‑occupation task in the prior month returned to that task 23.6 % of the time, compared with 8.4 % for comparable workers with no prior use. This recurrence rate exceeds that for within‑occupation and general tasks, indicating that cross‑occupation assistance is being integrated into ongoing workflows.

High‑Recurrence Cross‑Occupation Tasks

Task Return Rate Next Month
Discussing goods or services with customers 54 %
Advertising or promotional writing 44 %
Creating marketing materials 37 %
Explaining financial information 15 %
Average across cross‑occupation tasks 18.5 %

The variation suggests that AI fits more naturally into certain recurring activities (e.g., marketing) than others (e.g., financial explanations), possibly due to workplace norms or perceived risk.

Implications for Job Design and Organizational Strategy

The findings imply that work design should be considered alongside AI tool access when organizations plan AI adoption. As employees experiment with out‑of‑role tasks and find AI useful, those tasks can become regular responsibilities, broadening the functional scope of a job without changing its title.

OpenAI’s earlier Work at the Frontier report documented the initial emergence of task crossover. This follow‑up demonstrates that some of those crossover activities solidify into ongoing workflows, offering an early view of AI‑driven job transformation.

Future Research Directions

OpenAI plans to continue monitoring these shifts to better understand:

  • How the division of labor evolves as AI capabilities expand.
  • The long‑term impact on workers, businesses, and the broader economy.
  • Which occupational categories are most susceptible to early AI‑enabled expansion.

For deeper insight, readers can explore:

Sources