Source-led article
OpenAI study finds 43.5% of workers use ChatGPT for tasks outside their own profession
A new analysis by OpenAI suggests that the boundaries of professional roles are blurring as workers increasingly turn to ChatGPT to handle tasks outside their official job descriptions. The company examined more than 800,000 work-related ChatGPT messages and found that 43.5 percent of job-specific queries involved tasks associated with another profession — a pattern OpenAI calls “task crossover.”
The findings, reported by The Decoder, are based on messages where users explicitly mentioned their line of work and then asked ChatGPT to complete a task. OpenAI classified the tasks using the U.S. Occupational Information Network (O*NET), a standard database that maps activities to job profiles. Common tasks such as writing, summarising and scheduling were excluded to focus on specialised work.
Task crossover: what the data shows
Marketing and engineering tasks crossed over most frequently. For example, a person working in sales might ask ChatGPT to draft a technical specification, or a product manager might request a marketing campaign outline. Workers also used the AI tool to review contracts, perform data analysis, and troubleshoot websites — work that traditionally sits with lawyers, analysts, or IT teams.
OpenAI’s analysis indicates that these crossover queries are not rare exceptions. Over two-fifths of job-specific requests pulled skills from a different professional category. The company interprets this as an early signal that job profiles are shifting in practice, even before job titles or formal descriptions catch up.
Most affected tasks and professions
The data shows that crossover is not uniform across all fields. Roles that involve heavy documentation, data processing, or technical troubleshooting see the most cross-departmental use. Engineering teams reported use of ChatGPT for marketing briefs; marketing professionals used it to generate basic code snippets or technical analyses.
Because O*NET categorises activities rather than titles, the analysis captures work-style changes that would be invisible in conventional employment surveys. A customer support agent writing a Python script or a finance executive composing SEO copy both count as task crossover.
Stronger effect at small companies
The trend is particularly pronounced at smaller businesses and startups. In organisations without dedicated legal, data science, or IT departments, employees are handling specialised work themselves with AI assistance. OpenAI notes that the effect is stronger where specialist teams are thinner or absent.
This is especially relevant for India’s vast small and medium enterprise (SME) sector. Many Indian startups and growing businesses cannot afford full-time lawyers, data analysts, or marketing specialists. Instead, founders and generalists use ChatGPT to produce contracts, analyse sales numbers, and even debug code. The analysis suggests this pattern is becoming structural rather than occasional.
Methodology and caveats
OpenAI’s methodology relies on self-identified job roles in ChatGPT conversations, which may not always be accurate or representative. The sample covers only users who explicitly stated their profession, so the true crossover rate across the entire user base could differ. Additionally, O*NET is a U.S. framework; its categories may not perfectly map to Indian job markets.
Despite these limitations, the scale — over 800,000 messages — gives the analysis weight. It is one of the largest publicly described looks at how knowledge workers are actually using generative AI on the job.
What this means for Indian readers
For Indian professionals and business owners, the study reinforces a trend visible on the ground: AI tools are becoming a substitute for costly specialist hires. The data cannot predict whether this leads to a net loss of specialist jobs, but it does suggest that job descriptions are evolving faster than HR departments can update them.
The finding is also a reminder for educators and policymakers under India’s AI mission to consider how AI literacy across domains — not just coding — should be integrated into vocational training. If marketing officers need to read code and engineers need to write copy, curricula may need to reflect that.
Datos clave
| Metric | Value |
|---|---|
| Messages analysed | Over 800,000 work-related ChatGPT messages |
| Task crossover rate | 5% of job-specific queries |
| Most crossover areas | Marketing and engineering tasks |
| Effect more pronounced at | Smaller companies (fewer specialists) |
The article from The Decoder cites OpenAI’s analysis but does not link directly to a public dataset or blog post. As of publication, OpenAI has not released the full study. Readers should treat the figures as indicative of a trend rather than precise measurement, pending further peer review or official release.