Google Study: AI Adoption Widespread but Lacks Depth
A Google report analyzing 15 million AI interactions indicates that while AI adoption is broad globally, its use remains largely superficial.

A new report from Google, based on an analysis of 15 million anonymized AI interactions, suggests that artificial intelligence is being adopted widely across the globe, but often in limited ways.
The study, titled ATLAS (Activity, Task, Environment, and Adoption Study), compiled data from over 150 countries and 140 languages. It found that 68% of professional categories are now using AI to some extent. However, AI is integrated into only about 20% of the typical tasks within those professions.
Google summarizes the current state of AI as having wide coverage but shallow depth. In most real-world scenarios, AI is utilized primarily as an assistive tool. For automation, most workplace use is concentrated in collaborative activities like brainstorming, research, drafting, and learning, with fully automated tasks accounting for less than 10% of interactions.
The report also highlights that non-routine cognitive tasks, such as analysis and creative design, represent 65% of workplace AI interactions, despite making up only 35% of overall economic tasks. Outside of work, AI use is even more prevalent, with over 86% of interactions occurring in personal contexts like managing household affairs, shopping research, and interacting with government services.
Overall, the findings suggest that while AI's reach is extensive, its deep integration into core work processes remains limited, positioning it mostly as a supportive technology rather than a fully autonomous one.