Building AI agents for the enterprise

Our September 2025 Anthropic Economic Index found that in the U.S., 40 percent of employees report using AI at work, up from 20 percent in 2023. While these are significant gains, a looming question remains: will AI produce lasting advantages or incremental gains that plateau within a quarter? 

The enterprises pulling ahead are doing something different. They're embedding agentic AI into how employees work, how processes run, and what products are possible. They're encoding institutional knowledge into systems that compound over time, and they're seeing returns that are difficult for competitors to replicate.

To help leaders chart this path, we put together a guide on the three pillars of enterprise AI transformation, with examples from organizations doing this work today: L'Oreal, Lyft, and Rakuten. The guide also covers how to bring these capabilities to every team in your organization with Claude Cowork, without requiring a custom build for each one.

In this guide, we share:

• How to overcome the "agentic thinking divide" and why some AI deployments compound while others stall
• Best practices for upskilling employees with AI that reflects how your organization actually works
• Examples of how organizations can compress information-dense processes without sacrificing human-in-the-loop judgement and expertise
• Guidance for building new product capabilities that generate revenue, not just reduce cost
• A six-month deployment framework for rolling out Claude Cowork 
• Customer stories from L'Oreal, Lyft, and Rakuten

The companies getting the biggest returns from AI are being deliberate about how they teach it to employees, where they apply it, and what they build next.

Check it out, here. 

Get started with Claude Cowork today.

‍