How is AI diffusion measured?
Mostly as a share of people or firms using AI in a period. Microsoft’s AI Economy Institute counts the share of people aged 15 to 64 who used a generative AI product: 18.8% worldwide in June 2026, about a point above the first quarter, led by the UAE at 73.3% and Singapore at 64.3%, with the Global North at 28.8% against 16.2% in the Global South (Microsoft On the Issues, 21 September 2026, read 2 November 2026). Usage counts show breadth, not depth: a person asking a chatbot once a month counts the same as a team whose workflow runs on an agent.
Why does AI diffusion matter more than the next model release?
Because value arrives when ordinary firms change how they work, not when a lab sets a record. Capabilities from a frontier model spread to cheaper and open models over time, so every competitor ends up with similar tools. The advantage sits in how deeply a company redesigns its processes, not in which model it licensed first. That depth is what AI transformation tries to measure inside one company.
What was the AI diffusion rule?
A US export-control framework. The Bureau of Industry and Security published the Framework for Artificial Intelligence Diffusion on 15 January 2025, tightening controls on advanced computing chips and adding a control on the weights of certain closed-weight models (Federal Register, read 2 November 2026). Commerce announced its rescission on 13 May 2025, two days before compliance was due, and promised a replacement rule (BIS press release, read 2 November 2026). For a buyer outside the US the lesson is lasting: governments treat compute as strategic, which feeds sovereign AI programs and shapes the AI infrastructure vendors can offer.
AI diffusion vs AI adoption: what is the difference?
Scale and owner. Diffusion describes an economy; adoption describes your company.
| AI diffusion | AI adoption | |
|---|---|---|
| Scope | Countries, sectors, populations | One company or team |
| Typical measure | Share of people or firms using AI | Workflows people rely on daily |
| Who tracks it | Economists, governments, investors | Executives and process owners |
| Question it answers | Is AI spreading, and where? | Is AI changing how we work? |
Read diffusion figures as context for your plan, not as a target: high national usage says your competitors have the tools, while AI adoption inside your own walls decides whether you use them better.