Anthropic is expanding its Labs organization as it looks to accelerate the development of experimental products built on the latest capabilities of its Claude AI models. The company said the move is designed to support faster iteration at the frontier of model performance while creating a clearer path to scaling successful experiments into reliable products.
The Labs expansion builds on an internal development approach that blends rapid prototyping, early user testing, and selective scaling. That process has already produced several of Anthropic’s most visible offerings. Claude Code grew from a research preview into a billion-dollar product within six months, while the Model Context Protocol has reached roughly 100 million monthly downloads, becoming a widely adopted standard for connecting AI systems to external tools and data. Other products developed through this model include Skills, Claude in Chrome, and agent-based workflows that extend Claude beyond traditional chat interfaces, such as recently launched Cowork AI agent.
Anthropic said Labs will focus on pushing Claude’s agentic capabilities and testing new interaction patterns before they are rolled into core offerings. The goal is to give product teams room to experiment with unpolished ideas while maintaining a separate track for mature, enterprise-ready deployments.
Leadership Changes Support Dual-Track Strategy
As part of the expansion, Mike Krieger, co-founder of Instagram and Anthropic’s Chief Product Officer for the past two years, will join Labs to work alongside co-founder Ben Mann. Krieger’s shift reflects a greater emphasis on hands-on experimentation within the product organization.
Ami Vora, who joined Anthropic at the end of 2025, will assume leadership of the company’s Product organization. In that role, she will work closely with Chief Technology Officer Rahul Patil to scale Claude-based products used by millions of customers. The leadership changes are intended to separate exploratory development from operational execution, allowing each group to move at an appropriate pace.
“The speed of advancement in AI demands a different approach to how we build, how we organize, and where we focus,” said Daniela Amodei, Anthropic’s president. She said Labs is designed to support discovery at the edge of Claude’s capabilities while ensuring that successful concepts can be scaled responsibly for enterprise and consumer use.
The Labs expansion follows Anthropic’s recent move into regulated and enterprise-focused domains, including the rollout of expanded Claude capabilities for healthcare and life sciences. That release introduced HIPAA-compliant connectors and tools to support clinical trials, regulatory workflows, and scientific research, underscoring how Anthropic is applying its experimental-to-scale product model beyond productivity use cases. Together, these efforts highlight the company’s strategy of incubating new Claude capabilities through Labs before deploying them into high-impact, production-ready solutions across industries.
