Google Updates AI & Economy ATLAS With New Workplace Data

Google updated its AI & Economy ATLAS on September 15, 2026, with new visualizations and research on how different occupations use AI around the world.

The company reports that arts, design and media occupations account for 19% of work-related AI use in India, 1.6 times the global average. In the United States, computer and mathematical occupations represent 30% of work-related AI use, twice the share elsewhere.

WHAT THE ATLAS SHOWS: Google's AI & Economy ATLAS explores how AI usage relates to economic activity, including differences among regions and occupations. Rather than presenting a single global adoption figure, it offers a way to compare the composition of work-related use. Observed activity should not be confused with measured productivity.

INDIA'S 19% FIGURE: Google reports that arts, design and media occupations account for 19% of work-related AI usage in India, about 1.6 times the global average. This does not mean 19% of Indian designers use AI. The percentage describes the occupational composition of recorded work-related activity.

THE US COMPUTER-RELATED SHARE: Computer and mathematical occupations represent 30% of work-related AI usage in the United States, roughly twice the share in the rest of the world. The figure does not tell us how frequently an individual programmer uses AI or whether software quality has improved.

COMPOSITION IS NOT ADOPTION: A high occupational share can reflect the distribution of recorded activity rather than widespread adoption within that occupation. Estimating adoption would require a different denominator, such as the number of workers or users in a particular field. Reports should label these measures precisely.

WHY OCCUPATIONS DIFFER: Summarization, coding, brainstorming, translation and research are relevant to different tasks. Designers may experiment with concepts while developers may use AI to explain or test code. These are illustrative possibilities, not specific activities established by the ATLAS percentages.

POSSIBLE DRIVERS OF REGIONAL DIFFERENCES: Industry structure, language availability, education, connectivity and employer policies may shape patterns of use. Regions with large software industries may show more development-related activity. However, the observed distribution does not by itself identify the causes.

LESSONS FOR ENTERPRISE ROLLOUTS: Giving every employee the same AI tool may not produce the same benefits. Organizations should identify recurring tasks in each function and evaluate where assistance is useful. Regional variation is a reason to study actual workflows rather than assume one universal adoption pattern.

AI IN DESIGN WORK: Potential applications include early ideation, research synthesis, copy alternatives and simple interface prototypes. Human expertise remains important for interpreting user research and evaluating accessibility. A high volume of AI use is not evidence of superior design quality.

AI IN SOFTWARE DEVELOPMENT: Developers may use assistants for code completion, tests and explanations of existing systems. Generated code can still contain errors and vulnerabilities. A high occupational share of AI usage does not eliminate the need for review, testing and secure engineering practices.

USAGE VERSUS PRODUCTIVITY: More interactions with AI do not necessarily yield equivalent gains in output. Some tasks may be accelerated, while others require substantial correction. Business evaluation should consider completion time, quality and employee effort alongside usage counts.

DATA LIMITATIONS: Activity recorded through online services may not represent every worker or region equally. Connectivity and differences in preferred AI products can affect what is observable. Readers should review the data collection method before generalizing findings to a different workforce.

WHAT TO WATCH: The ATLAS is useful for asking where and how work-related AI activity is concentrated. Future research will be most informative if it can connect usage patterns to skills, job quality and outcomes. Occupational composition, adoption and productivity remain distinct concepts.

These figures describe the occupational composition of AI usage, not the percentage of workers in each occupation using AI. Regional and occupational context matters when interpreting adoption patterns.