Glean Survey Finds AI Saves 11 Hours a Week as Only 13% See Performance Gains
Updated
Updated · Financial Times · Jun 10
Glean Survey Finds AI Saves 11 Hours a Week as Only 13% See Performance Gains
3 articles · Updated · Financial Times · Jun 10
Summary
A survey of 6,000 digital workers found AI saves 11 hours a week, yet just 13% report any improvement in company performance.
Workers spend 6.4 hours of those gains on “botsitting” — feeding AI context, checking outputs, rerunning prompts and fixing mistakes.
Nearly 80% juggle multiple AI tools each week, and 60% repeat the same queries across platforms, creating a “toggle tax” that eats into efficiency.
A third of workers also downplay their AI use, underscoring how “workplace theatre” and legacy structures can blunt gains until companies reorganize teams, data and workflows around AI.
With AI investments nearing $700 billion, are we seeing history's largest capital misallocation or just a temporary productivity lag?
As AI decimates entry-level jobs, is the traditional college-to-career pipeline now permanently broken for the next generation?
The AI Productivity Paradox: Why 2025’s Rapid Adoption Hasn’t Delivered Big Returns—Yet
Overview
As of early 2026, AI adoption is accelerating rapidly across industries, especially in the Global North, where nearly a quarter of the working-age population uses AI tools—almost double the rate in the Global South. Despite this surge, the measurable impact of AI on productivity and profits remains modest, creating a paradox for businesses. This rapid uptake is mainly driven by high expectations of future benefits, but so far, the tangible results have not matched the hype. The report highlights this disconnect and explores why widespread AI integration has yet to deliver the significant gains many anticipated.