OpenAI Found That AI Is Blurring Career Boundaries: Workers Are Using LLMs To Do Other People’s Tasks

Workers are increasingly turning to ChatGPT to complete tasks that traditionally belonged to specialists in other professions, according to new research from OpenAI.

About 44% of occupation-specific ChatGPT requests involved work typically associated with another profession after excluding generic tasks such as drafting emails or scheduling meetings, Axios reported Monday, citing exclusive OpenAI research based on more than 800,000 work-related messages from U.S. business users.

"The boundaries between jobs are likely already becoming more flexible due to AI," Chatterji told the outlet, adding that workers are increasingly using ChatGPT to complete assignments outside their traditional roles.

The research found employees used ChatGPT for a wide range of specialist work, including creating marketing materials, troubleshooting software, performing financial calculations and explaining regulations. These are tasks that have typically been associated with dedicated marketing, engineering, finance or legal professionals.

Customer service employees, designers and human resources professionals recorded the highest levels of crossover work. Roughly three-quarters of occupation-specific prompts from those groups involved tasks more commonly linked to another profession.

OpenAI cautioned that the research does not establish whether AI is creating entirely new cross-functional responsibilities or simply helping employees perform duties that were already part of their existing roles. The researchers also said the analysis does not measure whether the requests resulted in higher productivity, better-quality work or different hiring decisions.

Smaller businesses appeared slightly more likely to rely on ChatGPT for work outside traditional job boundaries. According to the OpenAI research, about 19% of work-related requests at the smallest businesses crossed occupational lines, compared with roughly 16% at larger organizations. The Axios report said this pattern is consistent with smaller firms using AI to fill expertise gaps when dedicated specialists are unavailable.

The research adds another data point to the broader discussion around how generative AI is changing workplace practices. Earlier this year, Microsoft's 2025 Work Trend Index found that employees are increasingly using AI to support knowledge work across departments while organizations build what Microsoft described as "frontier firms" that combine human employees with AI agents. The report said many businesses are redesigning workflows around AI rather than limiting the technology to specific teams.

A separate report released by McKinsey & Company in June also found that organizations are moving beyond pilot AI projects and increasingly embedding generative AI into day-to-day operations, although adoption levels and measurable business results continue to vary widely across industries.

The OpenAI analysis focused only on business users in the United States and excluded broad workplace activities that could apply to almost any profession, such as writing emails or organizing calendars. Instead, researchers examined occupation-specific requests to better understand how employees were using ChatGPT beyond routine administrative work.

Rather than measuring employment trends, the study examined the types of work employees asked ChatGPT to perform. As a result, the research does not conclude whether AI is reducing jobs or changing workforce size. Instead, it documents how workers are applying AI across professional boundaries during their daily work.

The report concludes that AI may be changing the work employees perform before it changes overall employment levels, reflecting a shift in how workers distribute specialized tasks across organizations.