Analyzing more than 800,000 work-related messages from U.S. ChatGPT users linked to ChatGPT Business accounts, researchers found that 16.8% of all work-related requests involved tasks traditionally associated with a different occupation. When routine, broadly shared activities such as writing emails or scheduling meetings were excluded, cross-occupation tasks accounted for 43.5% of occupation-specific messages. The study examined 8 occupational groups: customer experience, design, engineering, finance, human resources, legal, marketing, and sales. Using the U.S. Department of Labor's O*NET occupational database, OpenAI compared users' reported roles with the tasks they asked ChatGPT to perform, describing this as "task crossover." Marketing and engineering tasks are the most likely to spread across occupations, while workers in design, sales, and human resources are more likely to use AI outside their traditional responsibilities. Cross-occupation AI use is more common in smaller organizations: among typical users, such tasks accounted for 18.9% of work-related messages in workspaces with 2 to 5 seats, compared with 16.3% in workspaces with 101 or more seats. The report emphasizes that findings do not indicate occupations are disappearing or that AI is replacing workers, but rather that AI is changing how work is divided, allowing employees to take on broader tasks while specialists remain essential for expert judgment, review, and accountability.