Anthropic経済指標についてClaudeに質問するAsk Claude about the Anthropic Economic Index

Anthropic has released a data connector that lets people query the Anthropic Economic Index directly inside Claude, turning what was previously a set of static reports and downloadable datasets into something users can interrogate through ordinary conversation. The move matters because it lowers the barrier to examining one of the more closely watched sources of evidence on how artificial intelligence is affecting work, allowing journalists, researchers, policymakers, and curious individuals to ask questions about labor markets and AI adoption without needing to write code or parse spreadsheets.

The Anthropic Economic Index is an ongoing research effort that Anthropic first introduced in early 2025. It draws on large volumes of anonymized and aggregated Claude conversations to estimate how AI is being used across different occupations and tasks. Rather than surveying workers, the index infers usage patterns by mapping conversations to occupational categories, frequently using task frameworks derived from the U.S. Department of Labor's O*NET database. The result is a picture of which kinds of work, from software development to writing to data analysis, show the heaviest AI engagement, and whether people appear to be using the technology to automate tasks outright or to augment their own efforts.

With the new connector, that underlying data becomes conversational. A user can ask Claude which occupations show the fastest growth in AI usage, how augmentation compares with automation in a given field, or how patterns have shifted across successive releases of the index, and Claude can respond by drawing on the connected dataset. This is a meaningful change in how the research is consumed. Previously, engaging with the index in depth generally required downloading published data or reading Anthropic's accompanying papers. Framing the data as something you can question directly is likely intended to make the findings more accessible to non-technical audiences and to encourage broader scrutiny of the numbers.

Technically, the feature fits within Anthropic's broader push around connectors, which give Claude structured access to external data sources and tools. Much of this ecosystem is built on the Model Context Protocol, an open standard Anthropic introduced in late 2024 to standardize how AI assistants connect to databases, applications, and other systems. MCP has since been adopted by a range of developers and companies as a common way to plug context into language models, and connectors like this one appear to be a natural extension of that approach, applying it to Anthropic's own published research rather than to a third-party service. Users should still treat conversational answers with appropriate caution, since a model summarizing a dataset can misstate figures or omit context, and verifying important claims against the source data remains sensible.

The release also sits within a wider industry conversation about AI's economic footprint. Other organizations have pursued related questions through different methods: labor economists and institutions have published studies estimating exposure of occupations to large language models, and companies including OpenAI and various academic groups have examined productivity effects in specific settings. Anthropic's index is distinctive because it is grounded in observed usage of a single widely used assistant rather than in hypothetical exposure scores, though that also means its view is shaped by who uses Claude and for what, which is a limitation the company has acknowledged. The data reflects a particular user base and cannot be read as a complete census of AI's role in the economy.

For Anthropic, making the index queryable through its own product serves several purposes at once. It demonstrates a practical application of connectors, reinforces the company's positioning as a contributor to public understanding of AI's societal effects, and keeps users engaged with Claude around a substantive topic. It also aligns with Anthropic's stated interest in informing policy debates, as accessible economic data can feed into discussions about workforce transitions, education, and regulation.

Whether the conversational format meaningfully changes how the index is used will depend on the reliability of the answers and the depth of the connected data. Still, the launch reflects a broader trend of turning datasets and documents into interactive resources accessible through natural language, and it offers a concrete example of how AI companies are applying their own tools to their own research outputs.