OpenAI Report Shows Employees Use AI to Broaden Job Scope Over Time
OpenAI released a report in its 'Work at the Frontier' series, analyzing over 1.5 million ChatGPT messages across approximately 6,200 employees over four months to study 'task crossover.' The findings show that employees using AI tools increasingly adopt and retain tasks outside their traditional job descriptions, transforming initial AI experimentation into permanent daily work routines. This research reveals that AI adoption does not just accelerate existing tasks but fundamentally reshapes workplace roles and organizational boundary lines. Organizations must rethink job design and workflow structure rather than merely treating AI as a software deployment project. The share of repeat non-traditional occupation tasks among employees' AI activities rose from 13.1% to 25.9% over four months, with tasks like customer discussions (54%) and marketing copywriting (44%) showing high monthly retention. Notably, when handling out-of-scope tasks, employees wrote shorter prompts and asked for fewer step-by-step explanations, but provided more background context and requested AI verification instead.
## BACKGROUND
'Task crossover' is a concept introduced by OpenAI to describe workplace behaviors where employees leverage generative AI tools like ChatGPT to perform tasks historically assigned to other specialist occupations. Understanding this phenomenon helps enterprise leaders measure the empirical productivity shifts and cross-functional capacity expansion caused by enterprise AI adoption.