Google opens ATLAS to measure global AI use

Google has made a new interactive version of AI & Economy ATLAS available to the public. The platform brings together millions of data points on how artificial intelligence is used across countries, professions and everyday activities.

The data shows that AI adoption is not progressing at the same pace everywhere. It is not concentrated in the same jobs, either.

India uses more AI in creative industries

In India, professions related to art, design and media account for 19% of work-related AI use. That is 1.6 times the global average.

The picture is different in the United States. Technical jobs lead there: computer and mathematical professions account for 30% of workplace AI use, twice the share recorded in the rest of the world.

The pattern also changes with the level of economic development:

• In OECD countries, computer, mathematical, business and financial professions stand out.
• In non-OECD countries, administrative work, creative industries and education are more prominent.
• Brazil and the United Arab Emirates record higher AI adoption than would be expected based on their GDP per capita.

The platform also lets you explore more specific uses. For example, equipment diagnosis and problem-solving in manual tasks account for 7% of workplace AI use in Brazil and Germany, compared with 4% in Japan. The global average falls between those figures.

Scientists save time, but build up a backlog

Google and Google DeepMind worked with MIT FutureTech on an analysis of AI use in science. The research combines data from 2,600 specialized models with a survey of more than 600 scientists in the United States and United Kingdom.

Nearly half of the scientists surveyed say they use some form of AI every day. They use both language models such as Gemini and systems designed for specific scientific tasks, although not for the same purposes:

• Language models are used across many fields and types of tasks.
• Specialized models are more common in health and life sciences.
• The latter are used especially to predict data, generate results and run simulations within a specific field.

Scientists say they save almost seven hours per week thanks to AI. That time can go toward research, analyzing results or preparing new work.

But saving time at one stage does not guarantee more discoveries. Researchers also spend time checking AI responses, and a list of hypotheses still waiting to be tested is growing. Bottlenecks appear at stages AI cannot speed up on its own, such as physical experiments and clinical validation.

What changes for you

ATLAS lets you see which professions use AI the most, for which tasks and in which countries. That helps distinguish real adoption from broad predictions about the future of work.

The main idea is simple: AI is already integrated into very different activities, from designing content to diagnosing equipment and supporting research. Its impact will increasingly depend on how work is reorganized around it, not only on what models can do.

Google presents ATLAS as a long-term research project. What to watch now is whether the time saved turns into more output, better decisions and new discoveries, or whether the processes that follow end up limiting those benefits.