OpenAI Research Finds AI Is Expanding Work Beyond Traditional Job Boundaries
An OpenAI analysis of more than 800,000 U.S. ChatGPT messages found that a large share of occupation-specific AI use involves tasks traditionally associated with another profession, offering an early view of how job responsibilities may be reorganizing.
Why it matters: Task crossover is a measurable signal that AI may change who performs work inside organizations before it changes formal job titles.
OpenAI Economic Research has published a new analysis examining how people use ChatGPT for work outside the traditional boundaries of their occupations.
The study reviewed more than 800,000 messages from U.S. users. OpenAI reported that 16.8% of work-related messages and 43.5% of occupation-specific messages involved tasks associated with another occupation. The company describes this pattern as task crossover.
Examples include marketers troubleshooting websites, salespeople exploring customer datasets and small-business owners handling basic financial, legal or communications work. OpenAI found that task crossover was particularly common in customer experience, design, human resources, legal and marketing roles.
The findings do not establish that entire jobs are disappearing. They indicate that the division of work inside organizations may change before job descriptions and official labor statistics catch up.
For employers, the practical question is increasingly how to redesign roles, approvals and training when employees can perform a broader range of tasks with AI support.