Anthropic just launched Claude Managed Agents, a product that makes building and deploying autonomous AI agents much easier. The launch comes as the company’s enterprise revenue soars, signalling a major push into the business AI space.

Making AI Agents Easier to Build and Deploy

Anthropic announced on Wednesday that it’s launching Claude Managed Agents, a platform aimed at reducing the complexity businesses face when building AI agents. The product supplies developers with ready-to-use infrastructure, allowing them to create autonomous AI systems without the heavy lifting involved in setting up the backend architecture.

Previously, companies had to deal with various software tools, memory management, and other technical hurdles to get AI agents running on their own. Claude Managed Agents wraps all those elements into a single package, so developers can focus on what the AI does instead of how it runs.

Riding a Wave of Enterprise Growth

Anthropic launched this product at a perfect time. The company revealed that its annualised recurring revenue has skyrocketed past $30 billion, tripling since December 2025. That’s a huge leap and shows just how fast the demand for AI tools in business is expanding.

Angela Jiang, who leads product for Anthropic’s Claude Platform, explains that most of this growth stems from the Claude Platform itself. This enterprise product lets developers connect to Anthropic’s AI models through an API, effectively tapping into the company’s powerful language models for a range of applications. Many developers have already created AI agents using this API, including tools like Claude Code, which assists with coding tasks.

But Jiang says there’s a gap between what their AI models can do and how businesses have been using them so far. Claude Managed Agents aims to close that gap by offering a streamlined way for companies to deploy a whole fleet of AI agents customised for their specific needs.

How Claude Managed Agents Works

At the heart of the new product is what Anthropic calls an “agent harness.” This is basically the software framework that surrounds an AI model, enabling it to act like an autonomous agent — performing tasks and making decisions on behalf of users.

The harness isn’t just software; it also includes a memory system that helps the agent remember previous interactions and a secure, sandboxed environment. This sandbox allows agents to launch software projects in a safe, isolated space, reducing risk while experimenting or running complex workflows.

Developers can also set permissions for their AI agents, controlling which tools the agents can access. The agents are capable of running autonomously for hours in the cloud, monitoring other agents, and adjusting their behaviour based on what’s going on.

Mastheads

Katelyn Lesse, head of engineering for the Claude Platform, points out that managing and scaling AI agents is a tough engineering challenge. Many of Anthropic’s customers had entire teams dedicated to building this kind of infrastructure before. Now, with Claude Managed Agents, those teams can focus on their core business instead of the technical headaches.

Real-World Use Case: Notion’s Client Onboarding

Anthropic demonstrated the product in action through a collaboration with Notion, a productivity startup. Eric Liu, a product manager at Notion, showed how Claude Managed Agents could automate client onboarding tasks within their platform.

In the demo, the AI agent handled a lengthy list of onboarding activities one after the other, freeing up human staff from tedious, repetitive work. Liu was able to monitor the process through a dashboard powered by Anthropic’s tools, highlighting how the product integrates smoothly into existing workflows.

Competition and Market Position

Anthropic isn’t alone in this race. OpenAI has been pushing its own agent platform called Frontier, aiming to capture the enterprise market with similar autonomous AI capabilities.

Both companies are racing to build out comprehensive solutions for businesses, especially as they eye public listings in the near future. Anthropic’s recent revenue surge suggests it’s gaining serious traction, but competition is fierce.

The enterprise AI market is evolving fast, with companies looking for tools that go beyond simple chatbots or language models. They want autonomous agents that can handle complex workflows, integrate with internal software, and operate securely at scale.

Implications for Businesses and Developers

Claude Managed Agents lowers the barrier for businesses wanting to experiment with AI agents. Instead of hiring specialised engineers to build and manage complex distributed systems, companies can tap into Anthropic’s ready-made infrastructure.

Smaller teams can now roll out AI automation faster, cutting costs and boosting innovation. For developers, it opens new possibilities to build smarter, more autonomous applications without reinventing the wheel.

The move toward autonomous AI agents is changing how businesses use artificial intelligence. Instead of just querying AI for answers, companies can deploy AI that acts independently — scheduling meetings, managing projects, or even monitoring other AI agents.

Still, managing permissions and security is a major challenge. Anthropic’s sandboxed environment and permission toggles aim to tackle these issues, but businesses will still need to be cautious about how much control they give to autonomous AI.

As AI agents get smarter and more common, concerns about oversight, accountability, and ethics will increase.

Related Articles

Anthropic’s Claude Managed Agents offer a fresh way for businesses to deploy autonomous AI, combining powerful models with pre-built infrastructure. With enterprise revenue climbing and competitors gearing up, the race to dominate AI agents is heating up — and companies wanting to automate work tasks have a new option to consider.

This article was created with AI assistance.