What and Why
AI is diffusing rapidly, and governments and multilateral institutions are directing billions toward digital infrastructure, skills programs, and labor policy. But they are acting with limited evidence on how AI adoption varies across countries and sectors.
The Global AI Adoption Index is a joint initiative of intergovernmental organizations, including the World Bank Group, IMF, and OECD, and leading AI companies — including Anthropic, Google, Microsoft, OpenAI, and OpenRouter. Unlike measures built on surveys or proxies, these organizations are working together on a harmonized method that leverages privacy-protected company data to track consumer chat and API adoption across countries, normalized against population, income, connectivity, and other national baselines.
These usage indicators would be supplemented with measures of conditions that shape adoption, including digital infrastructure, labor markets, and languages, drawing on complementary data and research from partners including Meta, LinkedIn, GitHub, and Ookla.
What We Are Building Towards
Establish the first harmonized, multi-company measurement system for global generative AI adoption, launched at the 2027 Global AI Summit.
Ensure the system is demand-driven and policy-relevant to project stakeholders.
Build coalition and infrastructure durable enough to persist beyond the 2027 publication as recurring statistical output.
The Challenge
Every major AI company holds a piece of this picture in its own usage logs, but no single company can see the whole.
Turning proprietary, competitively sensitive data from rival companies into one shared, comparable public good has not been attempted at this scale before.
DDP's Role
The Development Data Partnership is a consortium of 12 international organizations and more than 30 tech companies. DDP has an 8-year track record of facilitating private data sharing for public good.
This project applies DDP's experience creating privacy-preserving aggregation pipelines, data-sharing agreements that protect commercial sensitivity, and a participatory governance structure.
Who Can Use It and How
Government agencies — benchmark national AI adoption and target digital-economy investment.
Intergovernmental organizations — direct technical assistance and lending toward measurable adoption gaps; evidence-informed AI policy and regulatory advisory.
National statistical offices — build AI activity into labor market, productivity, and national accounts statistics.
Civil society and researchers — leverage a common foundational dataset for advocacy and research.
Companies — evaluate market, localization, and infrastructure priorities.
What We Aim to Measure as Quarterly Indices
Consumer chat intensity and diffusion — quarterly measures of adoption, by country
Consumer chat — harmonized sampling and taxonomy approaches across companies for analyzing how people use AI, by country
API token intensity — quarterly measure of use, by country
API users and use metrics — harmonized approach across companies for understanding model choices and industries (tbc)
Path to Publication
The Coalition
International Organizations
Publication Oversight - Chief Statisticians: World Bank Group (Haishan Fu) · IMF (Bert Kroese) · OECD (Steve MacFeely)
Contributing DDP Members:
Tech Company Commitment Signatories
In Discussion
Additional companies operating in different countries are in active discussion for participation.
Additional Components
US AI Adoption Index
State-level indicators launching alongside the global index, informed by civil society stakeholder consultations led by the Partnership on AI.
Supplementary datasets
A research program with LinkedIn, Meta, Ookla, and others building complementary datasets beyond chat and API data.
Low-resource language index
Accounting for new open-source datasets and audio/voice data, not just web crawl.