Anthropic Labs: Incubating Experimental AI Products

Anthropic has expanded Labs, a dedicated team designed to incubate experimental products at the frontier of Claude's capabilities. This organizational shift allows the company to separate the discovery of experimental tools from the scaling of stable, enterprise-ready products.

Organizational Structure and Leadership

Anthropic has restructured its product organization to support two distinct motions: experimental discovery and product scaling.

• Anthropic Labs: This team focuses on tinkering, experimenting, and testing unpolished versions of new capabilities with early users to identify viable products. Mike Krieger, co-founder of Instagram and former Chief Product Officer at Anthropic, joins Labs to build alongside Ben Mann.
• Product Organization: This team focuses on scaling the Claude experiences that millions of users rely on. Ami Vora, who joined Anthropic in late 2025, leads the Product organization in partnership with CTO Rahul Patil.

The Labs Approach to Product Development

Anthropic utilizes a specific motion for developing AI products, moving from research previews to scaled releases. This approach has already resulted in several key releases:

• Claude Code: Transitioned from a research preview to a billion-dollar product within six months.
• Model Context Protocol (MCP): Now an industry standard for connecting AI to tools and data, reaching 100 million monthly downloads.
• Additional Previews: Recent launches include Skills, Claude in Chrome, and Cowork, the latter of which brings Claude's agentic capabilities to the desktop as a research preview.

Strategic Rationale

According to President Daniela Amodei, the rapid advancement of AI requires a different organizational structure to "break the mold and explore." The Labs team provides the necessary environment to explore the frontier of Claude's capabilities while the broader product organization ensures these innovations are scaled responsibly for enterprise customers and a growing user base.