Google launched an open-access data tool for its AI & Economy ATLAS project on Tuesday, publishing workplace usage figures across global industries and regions. The release incorporates collaborative research conducted with Google DeepMind and MIT FutureTech that analyzes how researchers apply artificial intelligence to scientific work.
United States workers in computer and mathematical roles account for 30 percent of work-related AI usage in the country, double the share observed in the rest of the world. India recorded higher adoption in creative sectors, where arts, design, and media roles generated 19 percent of workplace AI usage, representing 1.6 times the global average.
OECD countries saw the highest AI adoption in technical, business, and financial operations. Non-OECD nations saw usage led by office and administrative support, education and library roles, and arts and media jobs. Google noted that while AI adoption generally tracks national income levels, Brazil and the United Arab Emirates showed higher adoption rates than their gross domestic product per capita would predict.
Scientific Workflows and Bottlenecks
Workers in Brazil and Germany directed 7 percent of their work-related AI activity toward equipment diagnostics and troubleshooting, which is 1.4 times the global average. Japanese workers directed 4 percent of their work AI usage toward those manual tasks.
Google, Google DeepMind, and MIT FutureTech also analyzed 2,600 specialized AI models and surveyed more than 600 scientists across the United States and the United Kingdom using an MIT FutureTech task taxonomy. Nearly half of the surveyed scientists reported using artificial intelligence tools daily, saving just below seven hours per week.
Scientists applied large language models like Gemini across diverse scientific fields, while reserving specialized models for life sciences, data predictions, and simulations. Researchers reported that these time savings created a growing backlog of untested hypotheses, with delays emerging around output validation, physical experimentation, and clinical testing. Google confirmed it will work with academic partners to expand ATLAS research.
