Anthropic announced on Wednesday the launch of Claude Managed Agents, a new product designed to simplify the process for businesses to build and deploy AI agents. This tool provides developers with out-of-the-box infrastructure to construct autonomous AI systems, streamlining a complex process that previously hindered the automation of work tasks.
This move positions Anthropic to capitalize further on its rapidly expanding enterprise business. The company reported on Tuesday that its annualized recurring revenue (ARR) has surpassed $30 billion, roughly three times higher than it was in December 2025. Anthropic, along with OpenAI, which also offers an agent platform called Frontier, is actively enhancing its enterprise offerings in anticipation of a potential public listing as early as this year.
According to Angela Jiang, Anthropic’s head of product for the Claude Platform, the majority of the company’s recent revenue growth stems from the Claude Platform, an enterprise product that enables developers to access Anthropic’s AI models via an API. Developers have been leveraging this API to deploy AI agents, such as Claude Code, within their workspaces.
Jiang emphasizes a notable gap between the capabilities of Anthropic’s models and their current business applications. The new Claude Managed Agents tool “enables any business to take the best-in-class infrastructure and deploy a fleet of Claude agents to do whatever work they need.”
Managed Agents will equip developers with an “agent harness,” which encompasses all the software infrastructure wrapping an AI model to facilitate agentic behavior—that is, taking actions on behalf of a user. Practically, a harness integrates software tools, a memory system, and other essential infrastructure. Agents created through Claude Managed Agents will also feature a built-in sandboxed environment, allowing them to initiate software projects securely. Furthermore, the product enables developers to create agents capable of autonomous operation for hours in the cloud, monitor the activities of other Claude agents, and manage permissions for accessing specific tools.
“When it comes to actually deploying and running agents at scale, that is a complex distributed-systems engineering problem,” states Katelyn Lesse, head of engineering for the Claude Platform. “Many customers we’ve spoken with previously required dedicated engineers to build and manage these systems at scale. By providing this capability out of the box, we allow those same engineers to focus on their business’s and product’s core competencies.”
In a demonstration shared with WIRED, AI productivity startup Notion illustrated its use of Managed Agents to power a client onboarding feature. Eric Liu, a Notion product manager, showcased how a long list of tasks within Notion could be off-loaded to a Claude Managed Agent, which systematically handled client onboarding steps. While the product in the demo ran within Notion, Liu also presented a dashboard on the Claude Platform for management.