OpenAI study: how ChatGPT expands the scope of work and blurs the lines between professions

In July 2026, OpenAI published a study on how artificial intelligence is changing the content of work. According to the company, ChatGPT users increasingly take on tasks from adjacent professions, and in doing so blur the familiar boundaries between job titles.

What the OpenAI study found

The OpenAI study records one key shift: AI does not so much replace workers as expand the set of tasks one person can perform. According to the company's observations, employees with access to ChatGPT step outside their formal roles: a marketer takes on rough code, a developer writes copy, an analyst prepares design mockups. The line between "my job" and "someone else's job" becomes fluid.

OpenAI describes this specifically as expanding rather than automation: the assistant lowers the barrier to entry for an adjacent skill, so people take on work they previously delegated to a colleague or a contractor. This continues OpenAI's line of research into how its models are actually used in real work scenarios.

Why professional boundaries are blurring

Professional boundaries are blurring because ChatGPT cuts the cost of learning a new skill to almost zero. When an assistant can help you write a SQL query, build a pivot table, or draft a legal clause in minutes, there is no longer a need to hand the task to a narrow specialist. OpenAI notes that this is exactly why a single employee covers more and more types of tasks within their role.

For companies, this changes the logic of hiring and team structure. Where a job was once defined by a narrow set of duties, the role now becomes broader, according to the logic of the OpenAI study, and the work becomes a combination of tasks that a human and AI perform together.

Who is affected

The shift mainly affects office and knowledge professions, where a task can be handed to a text-based assistant: marketing, development, analytics, support, management. In these roles, ChatGPT lets one person handle tasks that once required a separate specialist, which is why the boundary between professions is most fluid there. OpenAI treats this effect as cutting across functions rather than tied to a single industry.

What it looks like in practice

The expansion of tasks is visible in everyday operations: a person with an assistant combines functions that were previously split among several specialists. A marketer fixes simple code without waiting for a developer; an engineer polishes copy and documentation on their own; a manager pulls together analytics without a dedicated analyst. The mechanism is the same each time: ChatGPT lowers the barrier to entry for an adjacent skill, and the task does not leave the team for outside hands.

OpenAI is careful to describe expansion, not full interchangeability: the assistant helps people take on an adjacent task, but it does not remove the need for expertise where depth is required. The study records a shift in the breadth of tasks a single worker handles, not the disappearance of specializations.

AI expands what people do at work:

ChatGPT users take on tasks from different roles and redefine the boundaries between professions.

— from the

OpenAI study "How AI is expanding what people do at work"

What this means

The takeaway from the OpenAI study is simple: AI redefines not jobs but the content of work itself. The winners are not those whom AI will replace, but those who use it to take on a wider range of tasks, from code to copy to analytics, within a single role. For workers, this means betting on breadth and the ability to manage an assistant, not only on narrow specialization; for companies, it means rebuilding job descriptions around work that a human and AI do as a pair.