Anthropic’s Latest Offering Targets the Challenging Aspects of Developing AI Agents

On Wednesday, Anthropic revealed a new product designed to simplify the process of creating and deploying AI agents for businesses. Named Claude Managed Agents, this tool provides developers with ready-to-use infrastructure to create autonomous AI systems, easing a previously complicated process that hindered task automation.

This initiative positions Anthropic to leverage its swiftly expanding enterprise sector. The company announced on Tuesday that its annualized recurring revenue has exceeded $30 billion, approximately three times its figure from December 2025. Both Anthropic and OpenAI, which has its own agent platform called Frontier, are competing to enhance their enterprise offerings as they aim for a public listing as early as this year.

Angela Jiang, Anthropic’s head of product for the Claude Platform, states that most of the company’s recent revenue growth is driven by Claude Platform, which allows developers to access the company’s AI models via an API. Developers have been utilizing Anthropic’s API to implement AI agents, including Claude Code, within their work environments.

According to Jiang, there exists a considerable gap between the capabilities of Anthropic’s models and their actual applications in businesses. The new tool “allows any organization to leverage world-class infrastructure and deploy a fleet of Claude agents to perform necessary tasks,” Jiang explains.

Managed Agents will provide developers with an agent harness, which encompasses all the software infrastructure that supports an AI model in functioning autonomously or taking actions on a user’s behalf. This harness includes various software tools, a memory system, and additional infrastructure. Agents created using Claude Managed Agent will also feature a built-in sandboxed environment for securely initiating software projects. Moreover, the product enables developers to build agents that can operate independently for extended periods in the cloud, observe the activities of other Claude agents, and adjust permissions for accessing specific tools.

Katelyn Lesse, head of engineering for the Claude Platform, mentions, “Deploying and managing agents at scale presents a complex distributed-systems engineering challenge.” She notes that many customers previously required numerous engineers to construct and maintain these systems at scale. With the new out-of-the-box solution, those engineers can now concentrate on their core business and product competencies.

In a demonstration shared with WIRED, productivity startup Notion illustrated its use of Managed Agents for a client onboarding feature. Eric Liu, a product manager at Notion, showcased how he could delegate a lengthy list of tasks within Notion to a Claude Managed Agent, which effectively processed the client onboarding tasks sequentially. While the demo product operates within Notion, Liu accessed a dashboard on the Claude Platform to monitor the agents’ activities and the tools they were utilizing.

Recently, Wall Street investors have expressed concerns regarding software stocks, particularly as Anthropic launches a broad array of enterprise solutions, which some fear could render traditional software-as-a-service companies obsolete. Regardless of whether this concern materializes, Managed Agents clearly indicate that Anthropic still has considerable progress to make before most enterprises fully adopt Claude.