OpenAI Signals Maps ChatGPT Adoption Across 165 Work Activities Worldwide

OpenAI has published a sweeping analysis drawn from its Signals platform, examining how ChatGPT is used across 165 distinct work activities in nearly 100 countries. The dataset captures real-world adoption across roles — from software engineering and content creation to financial analysis and customer support — and tracks how intensity and frequency evolve over time.

Usage patterns shift toward higher-value tasks

Early adoption clustered around drafting, summarization, and coding assistance. Recent months show a measurable migration toward strategic work: research synthesis, decision support, and multi-step problem solving. This progression suggests organizations are moving beyond experimentation into production-grade workflows that require sustained model availability and low-latency inference.

Geographically, North America and Western Europe lead in absolute volume, while parts of Asia-Pacific and the Middle East show the steepest growth curves. The data also highlights sector-specific fingerprints — legal and finance lean heavily on document review, while marketing and design favor creative iteration.

Why it matters for GPU / AI infrastructure: As ChatGPT embeds deeper into core business processes, the underlying compute profile shifts from bursty, experimental traffic to predictable, high-throughput inference workloads. Cloud GPU providers must plan for sustained token throughput, regional capacity near user populations, and SLA-grade availability — not just peak benchmark performance.