What happened
Published 2026-06-26, this is the third edition of Anthropic's Economic Index series — the first to incorporate hourly-level telemetry and a linked survey (the Anthropic Economic Index Survey, launched April 2026). The report's central finding is that Claude usage now mirrors economic life in granular, measurable ways: personal use spikes from ~35% on weekdays to ~50% on weekends; tax queries surged before the April 15 US filing deadline; sleep-advice requests peak at 5 a.m. Chapter 2 shows that output type varies sharply by product, and that compute consumption rises with the estimated value of artefacts produced. Chapter 3 — the first results from the linked survey — finds that users who interact with Claude most autonomously (e.g., via API automation) are simultaneously the most optimistic about AI's impact on their pay, job security, and sense of meaning. The report also discloses a major methodological shift: chat transcripts no longer capture most Claude usage, which has shifted toward long-running agentic sessions via Claude Code and Cowork.
Why it matters
This is the most granular public window into how a frontier AI lab's products are actually being used at economic scale — and the survey data directly challenge the conventional narrative that heavy AI users fear job displacement. Executives designing AI workforce strategies should engage with the finding that high-automation users are the most optimistic cohort.
Action needed
Incorporate the survey finding — that employees using AI most autonomously report the highest optimism about pay and job security — into internal change-management communications and AI adoption roadmaps to counter workforce anxiety narratives.