Anthropic Says 35% Expect AI to Do Most Work Within 12 Months
Updated
Updated · Anthropic · Jun 25
Anthropic Says 35% Expect AI to Do Most Work Within 12 Months
2 articles · Updated · Anthropic · Jun 25
Summary
About 35% of surveyed Claude users said AI will be able to do most of their work within a year, in Anthropic’s latest Economic Index report linking roughly 9,700 responses to usage data.
The report pairs that survey with upgraded telemetry—hourly sampling, output classification and separate chat, Cowork and API breakdowns—to track how Claude use is shifting toward longer-running agentic tasks.
Usage patterns mirrored daily life and deadlines: personal conversations rose from about 35% on weekdays to just under 50% on weekends, recipe requests were 2.3 times more frequent at 6 p.m., and US tax queries jumped 8-fold before April 15.
Anthropic said heavier automation users were the most optimistic about AI’s effects on pay, job security and meaning, even as early-career workers reported the greatest concern about displacement.
Across work sessions, higher-wage tasks consumed more compute and often involved more user engagement rather than less, which Anthropic said points more to labor augmentation than outright replacement.
As workers hand over more tasks to AI, why does their job security optimism paradoxically increase?
As AI spending soars, how can businesses distinguish real productivity gains from a costly technological illusion?
With AI automating entry-level jobs, is the traditional career ladder for young professionals becoming a thing of the past?
35% of Users Predict AI Will Do Most of Their Work by 2027: Productivity Gains, Job Anxiety, and Widening Inequality
Overview
As of June 2026, AI is being rapidly integrated across industries, with user expectations for its future capabilities reaching new heights. Anthropic’s ongoing surveys reveal that over 35% of users believe AI will handle most of their work within a year, reflecting strong optimism. These surveys show that users often credit AI with a larger share of their work than external measures suggest, highlighting a gap between perception and observed data. This high level of confidence in AI’s potential marks a significant shift in how people view its role in transforming professional life.