Anthropic Launches Claude Science — AI Work Environment for Scientists in Beta

Anthropic, a developer of the Claude AI models family, on June 30, 2026, unveiled Claude Science—an application that unifies scattered researcher tools into a single workspace and allows AI agents to take on a significant portion of scientific work. The product is available in open beta status; according to the company, this is its most substantial step into the scientific sphere to date.

What Claude Science Can Do

Claude Science is conceived as a unified researcher's desktop. Instead of switching between dozens of scattered tools—from scientific literature search to data analysis—a scientist gets a single environment where AI agents based on Claude models can independently perform a significant portion of routine tasks.

• Launch date—June 30, 2026
• Product status—open beta
• Claude Science unifies scattered researcher tools in a single application
• AI agents based on Claude take on a significant portion of research work
• Anthropic calls this its largest step into the lab to date

Why This Matters for Researchers

Scientific work abounds with routine yet labor-intensive operations: literature searches, data systematization, preliminary analysis of results. These are precisely the kinds of tasks that Anthropic proposes delegating to AI agents within Claude Science, leaving researchers with more space for substantive work. The fact that the company calls launching its most substantial step into science shows that Anthropic views laboratories as one of the key applications of agentic AI, on par with other professional spheres.

Context: Agentic AI Goes Beyond the Chat Window

Recently, leading AI laboratories have been increasingly transforming models from simple chat interfaces into agents capable of performing multi-step tasks independently, rather than simply answering individual questions. Claude Science continues this logic: instead of limiting itself to responding to researcher queries, the system should act within their workflow—gathering data, preparing draft analyses, and taking on other steps that previously required manual researcher work from start to finish.

The choice of science as the platform for such a product is no accident: research work consists of many clearly formalizable stages where it's easy to measure whether an agent completed the task or not. This makes the laboratory a convenient testing ground for checking how far AI agent autonomy can be pushed before transferring such solutions to other professional spheres.

What This Means

The launch of Claude Science confirms that for Anthropic, AI agents are not merely a chatbot with extended capabilities, but a tool that embeds itself in a professional's real workflow and takes on part of their daily tasks. If the beta meets expectations, science could become one of the first spheres where agentic AI meaningfully accelerates the routine part of research work.

Frequently Asked Questions

What is Claude Science?

It's an application from Anthropic that gathers scattered researcher tools in one place and allows AI agents based on Claude models to perform a significant portion of scientific work.

When did Claude Science become available?

Anthropic announced the launch on June 30, 2026; the product operates in open beta status.