There is widespread agreement that AI has significant potential to transform the global economy and the way we work. But the results – what this means for work, people’s lives and the economy as a whole – are neither automatic nor guaranteed. A lot has to happen. To get there, we must work together as a society to positively shape the impact of AI on our lives, our work and our economy. For this collaborative work to be effective, it is critical to have a comprehensive understanding of how AI is being adopted and used in business. Society needs empirical insights and evidence-based research to make decisions, initiatives and actions.
To help, Google is launching the first edition of the AI & Economy ATLAS (Activity, Task, Landscape, and Adoption Study), an ongoing, large-scale, de-identified study of how people use Google’s AI products and tools. ATLAS’s first dataset (v1.0) is based on 15 million aggregated and anonymized human-AI interactions across the Gemini App, AI Mode and Gemini API, collectively used by more than 1 billion people monthly. ATLAS v1.0 insights span more than 150 countries, 140 languages, 800 professions and 4,000 tasks. ATLAS provides the most comprehensive look yet into how real people use AI at scale.
ATLAS sheds light on how people use Google’s AI tools for various tasks at work and in everyday life. The ATLAS v1.0 report provides an initial overview of a rapidly changing landscape: AI capabilities are advancing, its use is evolving, and tools to monitor its impact on the economy are still being developed.
Table of Contents
What do we learn from ATLAS v1.0?
Here are a few of the most interesting observations so far:
- The use of AI in the workplace is broad but superficial: Job acceptance extends across all industries and also to 68% of all occupations, which together account for 90% of total U.S. employment. However, within jobs, people use AI selectively: In a typical job, AI is only used for around 21% of tasks.
- At work, most AI usage focuses on collaboration and task assistance, and automating tasks is uncommon so far: ATLAS data shows that the vast majority of AI interactions in the workplace are focused on collaborative purposes such as ideation, strategy, information retrieval and learning. Tasks such as creative design and hypothesis testing (categorized as “non-routine cognitive” in ATLAS) appear much more frequently in AI work interactions than in business as a whole (65% vs. 35%). Less than 10% of these interactions involve tasks being fully automated.
- The use of AI is not limited to office workers, but also supports employees in predominantly physical and manual activities with related tasks: The use of AI in the workplace is not limited to jobs traditionally considered knowledge work. Although not as widespread, workers in manual and technical jobs (e.g., auto technicians, industrial mechanics) are using conversational AI as a live collaborator for real-time diagnostics, troubleshooting, and on-the-fly learning. When employees in these areas use our AI tools, they are twice as likely to use multimodal AI (i.e., use AI to create images or videos). For example, automotive technicians and industrial mechanics use AI to interpret complex test results, debug electrical wiring and check machines for wear.
- AI delivers value at home that may not be reflected in standard economic metrics, particularly in high-friction administrative tasks: Over 86% of interactions with AI tools in ATLAS occur outside of work. People are using AI in new and interesting ways that are not captured in standard economic metrics, including productive household activities (e.g., purchasing research, helping with the use of equipment and tools) and high-friction administrative tasks (e.g., navigating government services such as taxes, licenses, and fines).
- Global AI adoption tracks GDP per capita, with notable exceptions: The use of AI has spread worldwide. ATLAS data shows AI usage in over 150 countries and territories, representing 99% of the world’s population. We also see this in the variety of languages used in ATLAS. English only accounts for about a third of global AI conversations, and users do not systematically abandon their native language for complex tasks. Upon closer inspection, AI usage per capita accurately reflects a country’s relative wealth level, raising concerns about a persistent digital divide. However, this is not a universal rule: some middle-income countries in South America and the Middle East are adopting AI at comparable rates to higher-income countries.
Here are some further insights:
https://blog.google/innovation-and-ai/technology/research/understanding-the-ai-economy/
