AI Redefines Jobs: A Transformation Underway

The Impact of AI on the Distribution of Professional Tasks

Artificial intelligence (AI) is radically transforming the way professional tasks are distributed among workers. An in-depth analysis of over 800,000 messages exchanged by ChatGPT users in the United States revealed that 16.8% of work-related messages and 43.5% of job-specific messages pertain to tasks traditionally associated with other professions. This indicates a significant shift in how roles are defined and executed.

Take, for example, a small business owner who, thanks to AI, can now write texts, review contracts, or even conduct basic financial analyses without the help of specialists. Similarly, a salesperson can use AI tools to analyze customer data, a task that would have previously required the intervention of an analyst. A marketer, on the other hand, can troubleshoot technical issues on a website without waiting for a developer's assistance. These examples illustrate how AI is changing not only work methods but also the distribution of tasks among different professionals.

The report titled Work at the Frontier: How AI is Expanding What People Do at Work delves deeply into these transformations. It introduces the concept of task crossover, where tasks historically associated with one profession appear in the AI usage of individuals from other professions. This phenomenon is analyzed within our AI Jobs Transition Framework, which suggests that many jobs are likely to undergo significant reorganization. This report is the first in a series titled Work at the Frontier, aimed at providing data-driven insights to guide work-related policies and practices.

The Extent of Task Crossover Across Different Sectors

Studies on AI and work often begin with a fixed list of tasks associated with a given profession, assessing whether AI models can perform them. However, our research shows that AI is also altering the distribution of tasks among professions.

To measure this crossover phenomenon, we first distinguished between profession-specific work and work common to multiple jobs. Certain activities, such as writing, summarizing, and planning, are too broadly shared among professions to be considered examples of crossover. We classified them as generic. For other messages, we checked whether the task fell within the user's profession or not. Among non-generic messages, 43.5% pertained to a different profession, providing an early glimpse into how AI might reshape task content before these changes are reflected in job descriptions or titles.

This finding suggests that a significant portion of work-related ChatGPT usage comes from users who are expanding their roles. This pattern is particularly pronounced across several professional groups. Once generic work is excluded, out-of-profession tasks represent:

• 77% of job-specific messages from customer experience workers
• 75% from designers
• 69% from human resources workers
• 56% from legal workers
• 53% from marketers

These professions "borrow" tasks from other roles, indicating an evolution in the division of labor. Some activities that previously required a transfer can now be performed by the person who first encounters the need.

The Tasks That Travel the Farthest

In examining task crossover, we found that some tasks travel farther than others. Marketing and engineering tasks, for example, frequently appear in messages from workers outside these fields.

Share of Tasks Within the Profession

Task crossover is not evenly distributed when we take a closer look at professions and specific tasks. Financial calculations and technical troubleshooting rank among the three most common tasks outside the seven other professional groups analyzed. Marketing work also travels widely: creating marketing materials appears in five other groups and is particularly notable among design users.

By analyzing broader task sets, two distinct directions of crossover emerge. Some jobs incorporate many tasks from other professions, while others provide tasks that appear across many different jobs.

Design illustrates the first pattern. About 35.2% of messages from designers involve work typically associated with another profession, while design tasks account for only 1.7% of messages from workers in other fields. Designers heavily rely on external tasks, but design work itself rarely appears elsewhere.

Engineering, on the other hand, is closer to the opposite. Only 18.5% of engineering messages involve tasks from other fields, but engineering tasks represent 7.4% of messages among workers in other professions. Engineering is a significant source of work that others take on, ranging from software troubleshooting to managing technical systems.

Marketing stands out in both directions. Marketers dedicate 24.3% of their messages to tasks associated with other professions, while marketing tasks account for 8.9% of messages among workers in other fields — the highest external share in the sample. Marketing workers combine tasks from multiple areas, and marketing work also spreads widely within the organization.

The Influence of Company Size on Task Crossover

The size and structure of a company play a crucial role in how AI modifies work. In a large company, employees may have access to specialized teams, established workflows, and internal services. In contrast, in a small organization, the worker closest to the problem is more likely to resolve it themselves rather than delegate.

Among average users, the share of out-of-profession tasks decreases from 18.9% for users in workspaces of 2 to 5 positions to 16.3% for users in workspaces of over 100 positions. However, among the most intensive users, we do not observe the same monotonic pattern.

One possible explanation is that moderate users in small organizations turn to AI when they encounter work that would otherwise require another function. Intensive users may have developed AI-supported workflows that resemble one another across organizations, or they use AI more intensively within their primary profession.

AI can be particularly useful as a generalist tool when specialized resources are scarce.

The Evolution of Job Descriptions

AI usage data like this serves as an indicator of how work is evolving. It allows us to see how AI enables workers to experiment with new combinations of activities before companies revise job descriptions or create new titles. In this sense, usage patterns can provide an early signal of professional change that conventional labor market statistics will only capture later.