Anthropic launched a connector on July 22 that lets any Claude user query the Anthropic Economic Index directly through a normal conversation. You open the connectors menu in claude.ai, enable the Index, and start asking questions. No technical skills required.
The Index itself is worth understanding before you use it. It is built from millions of anonymised conversations on Claude, analysed to identify what kinds of work tasks people are actually doing with AI, by occupation and location. The data covers every US state, dozens of countries, and hundreds of job types. Crucially, it tracks whether AI is being used to assist humans or to replace them entirely.
The headline number from the latest data: nearly half of all jobs tracked have seen AI used for at least 25% of their associated tasks, up from 36% when the Index launched in early 2025. That is a significant jump in just over a year.
Not all sectors are moving at the same speed. Customer service workflows are already heavily automated through enterprise API use. Sales and business outreach workflows roughly doubled between late 2025 and early 2026. These are not slow-moving trends.
The Index also draws a useful distinction between two modes of AI use. Augmentation is when AI helps a person do their job better, they stay in the loop, check the output, and make the final call. Automation is when the AI does the task from start to finish, without human involvement. On Claude's consumer product, augmentation is slightly ahead. On the enterprise API used by companies building workflows, automation dominates at around 75%.
That gap matters. When companies build automated workflows into their operations, they are making structural decisions, not experimenting. Those decisions tend to be harder to reverse and they compound over time.
Another finding worth noting: the average hourly wage of Claude users has been drifting slightly downward, from $49 to $48, as more non-technical workers adopt the tool. Early users were mostly software developers. The user base is broadening, which means AI adoption is no longer a story only about technical jobs.
A survey of 81,000 Claude users found that people report large productivity gains but also express genuine concern about job displacement. Those concerns are highest among early-career workers and in roles where the Index shows the most AI activity. That combination, high productivity gains plus anxiety about the future, is the real picture on the ground right now.
For any manager or director trying to make sense of where their industry is heading, this tool is worth ten minutes of your time. You will get grounded, data-based answers instead of predictions. Ask what tasks people in your sector use Claude for. Ask whether the trend in your field is toward assistance or replacement. Then decide what that means for how you plan the next 12 months.