OpenAI says a growing share of workers are using ChatGPT to take on work that would normally sit outside their own profession. The company describes the pattern as "task crossover", a sign that AI tools are changing how workplace responsibilities are divided.
The finding comes from an analysis of more than 800,000 work-related ChatGPT messages. Among job-specific queries in that set, 43.5 percent involved another profession, according to OpenAI.
What OpenAI Means by Task Crossover
Task crossover refers to cases where a person uses ChatGPT for work associated with a different job profile. In practical terms, that can mean a worker asking for help with a task that once would have been assigned to a specialist.
OpenAI says users handled work such as contract reviews, data analysis, and website troubleshooting. Those examples point to a broader pattern: people are not only using AI for general productivity, but also for job-specific work that crosses professional boundaries.
The company classified tasks using the U.S. occupational database O*NET, which maps work activities to standard job profiles. It also excluded common tasks such as writing, summarizing, and scheduling, focusing instead on more profession-linked activity.
Marketing and Engineering Cross Over Most Often
OpenAI found that marketing and engineering tasks crossed over most often. The source does not say which workers were taking on those tasks, but it does show that these two areas appeared frequently when users worked outside their own occupational lane.
That matters because both categories often touch other parts of a company. Marketing work can connect to sales, product communication, and customer-facing materials. Engineering work can connect to websites, technical fixes, and systems that other teams depend on.
The analysis suggests ChatGPT is being used as a bridge into those areas. It does not mean every worker becomes a specialist. It does suggest that more people may be attempting specialist-adjacent work with AI assistance before handing it off, escalating it, or deciding whether expert help is needed.
Why Smaller Companies May Feel the Shift First
OpenAI says the effect is stronger at smaller companies, where dedicated specialist teams are less common. That makes the pattern easier to understand. When a company has fewer people, workers often already carry broader responsibilities, and ChatGPT can make it easier to attempt tasks beyond a formal role.
For a smaller organization, task crossover may look like practical problem solving. A person who is not a lawyer may use ChatGPT to review a contract. Someone outside analytics may use it for data analysis. A non-engineer may ask for help with website troubleshooting.
The source does not claim these workers are replacing specialists. It shows that they are using ChatGPT to engage with work that has traditionally belonged to other professions. That distinction is important: the data points to shifting behavior, not a complete rewrite of company structure.
Job Profiles May Be Changing Before Titles Do
OpenAI sees the usage data as an early signal that job profiles are shifting, even before job titles or descriptions catch up. In other words, the work people actually do may be changing faster than the formal language companies use to define roles.
This is one of the clearest implications of task crossover. If workers are repeatedly using ChatGPT to perform job-specific tasks from other professions, then the boundary between roles may become less rigid in day-to-day operations.
That does not require a company to rename jobs immediately. It does suggest that the practical scope of a job can expand quietly through tool use. A role may still have the same title while the person holding it handles a wider range of work.
What the Data Does and Does Not Show
The analysis is specific in several ways. It covers more than 800,000 work-related ChatGPT messages, identifies 43.5 percent of job-specific queries as involving another profession, and excludes common tasks such as writing, summarizing, and scheduling.
Those boundaries help clarify the finding. OpenAI was not simply counting everyday office help. It was looking at job-specific work and comparing those tasks with standard job profiles through O*NET.
At the same time, the source does not say whether the resulting work was accurate, approved, or used in final form. It also does not say how companies responded internally. The strongest supported conclusion is narrower and still significant: more workers are using ChatGPT to step into tasks associated with other professions.
For employers and workers, that makes task crossover a useful concept to watch. It gives a name to a workplace behavior that may already be happening before policies, job descriptions, or team structures formally recognize it.