Google study of 15M AI chats finds adoption broad but shallow

Google Alphabet Inc. ↘ 0.96% on Wednesday released what it calls the most comprehensive look to date at how people use AI in their daily lives and work, drawing on 15 million de-identified interactions with its Gemini products. The central finding: AI adoption is remarkably broad but shallow, with workers using the tools for assistance rather than full task automation.

Broad Reach, Thin Penetration

The study, called ATLAS (Activity, Task, Landscape, and Adoption Study), analyzed interactions across the Gemini app, AI Mode in Search, and the Gemini API during a two-week window in April 2026. Google reports that AI use appears in 68% of detailed occupations — jobs that together cover roughly 90% of U.S. employment — yet within a typical role, workers reach for it on only about a fifth of their tasks.Fox Business+2

Fewer than 10% of workplace interactions fully automate a task. Instead, most usage focuses on collaborative activities: ideation, research, drafting, and learning. Work classified as "non-routine cognitive," such as analysis and creative design, accounts for 65% of workplace AI interactions, compared with 35% of tasks in the economy overall.LinkedIn+1

Beyond Knowledge Work

The findings extend past traditional white-collar roles. Workers in manual and technical trades — including auto technicians and industrial mechanics — use conversational AI as a real-time diagnostic aid and are twice as likely to use multimodal features such as image input. The study spans more than 150 countries, 140 languages, 800 occupations, and 4,000 tasks.Unite.AI+1

More than 86% of all interactions happen outside formal work, in household administration, shopping research, and navigating government services like taxes and licensing.LinkedIn+1

Context and Caveats

Google developed the study through its AI and Economy Research Program, working with outside academics including MIT's David Autor and Cambridge's Diane Coyle. However, the dataset excludes Google's paid enterprise products — Gemini Enterprise, Workspace, and Google Cloud — meaning it captures consumer and prosumer behavior rather than deep corporate deployments.Unite.AI

The report arrives as Google's AI capital spending outpaces its cash flow and as vendors across the industry race to sell automation and agentic AI as the next enterprise upgrade. A first-party dataset showing automation remains rare offers what Unite.AI called "a useful hedge against expectations it cannot yet meet".Unite.AI