Anthropic: Economic Index reveals daily rhythms and patterns of AI usage
The third Anthropic Economic Index report analyzes when and how people use Claude — from morning news to late-night recipes. Claude Code's autonomy score is 0.37 points higher than chat, and conversations in high-paying occupations consume 2.5x more tokens.
This article was generated using artificial intelligence from primary sources.
What is the third Anthropic Economic Index?
The Anthropic Economic Index — a longitudinal research project tracking real patterns of Claude usage in work and everyday contexts — published its third report in June 2026. The key finding: as many as 93% of Claude conversations produce recognizable, concrete outputs, suggesting that users are primarily using AI as a productive tool rather than for experimentation.
When and how do users reach for AI?
The data reveals pronounced temporal rhythms. News-related queries peak at 7 am, while recipe queries rise 2.3x in the evening around 6 pm. The most illustrative example of seasonality: tax queries jump +8x in the week before the April 15 deadline — an almost identical pattern to search engines.
Claude Code vs chat: what does the autonomy score reveal?
The report introduces a comparison by autonomy score — a measure (scale 1-5) of how independently AI drives a task without constant user guidance. Claude Code sessions record an autonomy score 0.37 points higher than standard chat, reflecting long, structured coding tasks as opposed to short interactive conversations.
High-paying occupations consume significantly more resources
Conversations from users in high-paying occupations consume on average 2.5x more tokens than the average — indicating more complex, multi-layered queries and longer sessions. A survey of 9,700 respondents shows that 60% expect significantly greater AI capabilities within the next 12 months. A gender pattern is also visible: women show more iterative working styles and a slightly lower share of fully automated tasks compared to male users.
The third report confirms that AI has entered everyday rhythms — not only as a productivity tool, but as a contextually aware assistant that follows the user’s daily schedule.
Frequently Asked Questions
- What does the autonomy score measure?
- The autonomy score (scale 1-5) measures how independently AI handles a task without human guidance. Claude Code sessions record an autonomy score 0.37 points higher than standard chat.
- What are the most pronounced temporal usage patterns?
- Recipe queries rise 2.3x at 6 pm, news queries peak at 7 am, and tax queries jump as much as 8x in the week before the April 15 deadline.