Task crossover: How AI is redistributing work across roles in SMEs

Summary

OpenAI reports findings from an analysis of over 800,000 work-related messages from U.S. ChatGPT users, showing that AI enables employees to perform tasks traditionally associated with other roles. This pattern, called task crossover, shows up in 43.5% of occupation-specific ChatGPT use and is particularly pronounced in smaller organisations where specialist resources are scarce. Read the full report.

Key Claims

• In 43.5% of work-related ChatGPT use, users tackle tasks outside their own occupation; the rate is higher in smaller workspaces (18.9% cross-occupational tasks in 2-5 person teams vs 16.3% in teams over 100).
• Customer experience (77%), design (75%), and HR workers (69%) show the highest cross-occupational task rates, suggesting these roles are most prone to internal task redistribution through AI.
• Marketing and engineering tasks travel furthest across the organisation: engineering tasks account for 7.4% of messages from workers outside engineering, and marketing work spreads even wider (8.9%), so both are candidates for company-wide AI rollout.
• Financial calculation and technology troubleshooting rank in the top three outside-occupation tasks across all seven occupation groups, showing where AI delivers value across nearly all roles.
• In small organisations, the worker nearest the problem is likelier to do it than delegate, so AI-as-generalist tool works best where specialist teams do not exist.

Quotes

• "Some activities, such as writing, summarizing, and scheduling, are shared too broadly across occupations to be evidence of crossover."
• "Marketing workers combine tasks from multiple domains, and marketing work also spreads widely across the organization."
• "AI may be especially useful as a generalist tool where specialist resources are scarce."