Workers Use ChatGPT for Tasks Outside Their Job Descriptions

Workers are increasingly using ChatGPT to handle tasks that traditionally fall outside their professional expertise, according to new research from OpenAI analyzing more than 800,000 work-related messages from U.S. business users.

The study found that 44% of occupation-specific ChatGPT requests involved work typically associated with another profession—after excluding generic tasks like email writing and meeting scheduling. Workers turned to the AI tool to create marketing materials, troubleshoot software, perform financial calculations, and explain regulations, all tasks normally handled by specialists.

Why it matters

This pattern suggests AI is reshaping the division of labor within organizations before those changes appear in official employment data. If sustained, this trend could fundamentally alter hiring practices and job design, as employees gain the ability to perform specialist work that previously required dedicated expertise.

Which professions cross boundaries most

Customer service workers showed the highest rate of crossover work, with 77% of their occupation-specific prompts involving tasks more commonly associated with other professions. Designers followed at 75%, and human resources professionals at 69%.

Engineers demonstrated the lowest crossover rate at 28%, while legal professionals landed in the middle at 56%. Marketing, sales, and finance workers all showed crossover rates between 40% and 53%.

Small businesses lead adoption

Among typical ChatGPT users, nearly 19% of work-related requests at the smallest businesses crossed occupational boundaries, compared to roughly 16% at larger firms. The research suggests this pattern reflects workers at smaller companies using AI to fill expertise gaps when specialists aren't readily available on staff.

"The boundaries between jobs are likely already becoming more flexible due to AI," OpenAI chief economist Ronnie Chatterji said.

Key limitations

The researchers acknowledged important constraints in their analysis. The study cannot determine whether AI is creating entirely new cross-occupation work or simply helping workers perform responsibilities they already held. It also doesn't measure whether these ChatGPT requests translate into higher productivity, comparable work quality, or different hiring decisions by employers.

The findings point to a broader conclusion: AI may change the nature of work people perform before it changes the number of people employed. This distinction matters for business leaders planning workforce strategies and policymakers tracking labor market transformations.

These details were first reported by Axios, which obtained exclusive access to the OpenAI research.