Anthropic launches Claude Managed Agents for business AI agents

In short: Anthropic launched Claude Managed Agents, a new product that helps businesses build and run AI agents more easily.

What happened

Anthropic announced Claude Managed Agents on Wednesday. The company says the product is designed for businesses that want to create AI agents, meaning AI tools that can carry out tasks on a user’s behalf, not just answer questions.

Anthropic says many companies have struggled to deploy agents because the supporting software is complicated. Claude Managed Agents provides that support out of the box, including what Anthropic calls an “agent harness” (the surrounding parts that help an AI system operate, like tools, memory, and controls).

The product includes a built-in “sandboxed” environment (like a locked testing room) so an agent can spin up software projects in a more secure setting. Anthropic also says agents can run on their own for hours in the cloud (remote computers), watch what other Claude agents are doing, and use adjustable permissions so companies can limit which tools an agent can access.

In a demo shared with WIRED, Notion showed a client onboarding feature powered by Claude Managed Agents. A Notion product manager assigned a long list of onboarding tasks, and the agent worked through them one by one. The manager also viewed a Claude dashboard that showed what the agents were doing and which tools they were using.

Why it matters

For regular people, this could affect how quickly workplaces automate routine tasks, like setting up new customers or organizing internal work. Anthropic is also pushing further into business software, as it reported annualized recurring revenue above $30 billion and competes with OpenAI, which has its own agent platform called Frontier.

Source: Wired