Google's AI & Economy ATLAS finds scientists save nearly 7 hours a week using AI

Google published new findings from its AI & Economy ATLAS research initiative on September 15, 2026, alongside a new interactive, open-access data tool, showing sharp regional and occupational differences in how AI is being adopted — plus a companion study finding scientists save nearly seven hours a week using AI in their work.

What's new

The post opens by asking "how is AI being used by people every day, including in offices, labs, and creative studios around the world?" and states that "insights from Google's AI & Economy ATLAS show how AI is actively reshaping global work."

Headline regional findings:

• India's creative sector uses AI at a notably higher rate than the rest of the world, with "arts, design, and media occupations making up 19% of work-related AI usage, 1.6 times the global average"
• The U.S. leads in technical adoption, with "computer and mathematical occupations accounting for 30% of work-related AI usage, double the share in the rest of the world"
• In Brazil and Germany, 7% of work AI usage goes toward real-time equipment diagnostics and troubleshooting (1.4 times the global average), compared with 4% in Japan
• Adoption generally tracks a country's income level, but Brazil and the UAE show higher adoption than GDP per capita would predict

Google also launched a new interactive, open-access tool for exploring ATLAS's underlying data by occupation, country, and use case — from electricians to purchasing managers.

Alongside the ATLAS update, Google, Google DeepMind, and MIT FutureTech published a companion study on scientific AI use, drawing on an analysis of 2,600 specialized AI models and a survey of more than 600 U.S. and U.K. scientists. Its findings: "nearly half of surveyed scientists use some form of AI every day," and scientists are "reporting saving almost 7 hours a week with AI." The study notes that gain creates a new bottleneck: faster idea generation is producing "a backlog of hypotheses" further down the research pipeline that hasn't sped up to match.

Context

AI & Economy ATLAS is Google's ongoing effort to quantify real-world AI usage and economic impact using aggregated product data rather than self-reported surveys alone, similar in spirit to Anthropic's Economic Index and OpenAI's own usage-pattern reports. This release adds a scientist-specific study built with MIT FutureTech's task taxonomy for scientific work, and ships a public interactive tool rather than a static report.

Why it matters

The report adds hard, large-scale usage data to a debate that has mostly run on anecdote: not just whether AI helps knowledge workers, but where the gains concentrate and where they stall. The scientist finding is the sharper point — AI is measurably compressing the idea-generation phase of research, but the study's own authors say that's now creating a backlog rather than a straight-line productivity win, because downstream validation and experimentation haven't sped up to match. That distinction matters for how research institutions, funders, and AI labs building science-focused tools think about where the next bottleneck actually is.