OpenAI: cross-role AI tasks jump to 25.9%

OpenAI's latest Work at the Frontier report builds on an earlier finding of 'task crossover' — workers using AI for activities historically associated with another occupation. The new question was whether those cross-occupation uses stick. Analyzing more than 1.5 million work-related ChatGPT messages from April through July 2026, OpenAI found that among roughly 6,200 workers observed consistently, cross-occupation tasks grew from 13.1% of occupation-specific AI activity in April to 25.9% in July, a pattern consistent with workers folding such tasks into ongoing workflows rather than trying them once.

A separate matched follow-up analysis supports that reading: workers returned to a cross-occupation task used the previous month 23.6% of the time, versus 8.4% use of the same task among comparable workers with no observed prior use. Recurrence is uneven across task types, ranging from 54% for discussing goods or services with customers to about 15% for explaining financial information, against an 18.5% average. OpenAI suggests this could reflect where AI fits naturally into recurring workflows, or differences in workplace norms and perceived consequences of error.

The takeaway OpenAI draws is that AI may reshape jobs before titles change, with a worker experimenting outside their traditional role, finding AI useful, and returning to that activity regularly. For organizations, that suggests work design may deserve a place alongside tool access in AI strategy. The patterns described are correlational and open to interpretation, so whether cross-occupation tasks ultimately become formal responsibilities — and how that reshapes the division of labor — remains an open question OpenAI says it will keep studying.