OpenAI's new Signals dataset maps ChatGPT usage at the country level, revealing how adoption patterns, query types, and behavioral trends diverge across global markets. This is not a marketing report. It is granular behavioral data showing where users have shifted from occasional curiosity to habitual, task-driven workflows.
The core finding is in the title: users are moving from asking to doing. That means fewer exploratory prompts, more delegated tasks, longer sessions, and repeat engagement on complex work. The data captures this shift with country-specific breakdowns, which is the part worth reading carefully. Adoption curves look different in Brazil versus Germany versus Japan, and the reasons why are embedded in the numbers.
What makes this worth reading in full is not the headline conclusion but the methodology behind the country-level segmentation and what OpenAI chose to measure. The categories of use, the behavioral proxies for 'doing versus asking', and the markets showing the steepest growth curves all carry implications for how AI products will be built and prioritized in the next 18 months.
[READ ORIGINAL →]