Anthropic has introduced a new feature called Claude Code Channels, allowing developers to interact with its AI coding agent through messaging platforms such as Telegram and Discord. The update expands how users can manage AI-driven workflows beyond the terminal.

The feature, currently in research preview, enables users to send messages, alerts, and automated events directly into an active Claude Code session. This allows the AI agent to respond to inputs even when the user is not actively working in a development environment.

Claude Code Channels function as a bridge between external platforms and the local AI session. Messages sent through supported platforms are delivered into the session, where the AI can process requests, execute tasks, and send responses back through the same channel.

Extending AI Beyond the Terminal

The integration reflects a broader shift toward persistent AI agents that operate continuously and respond to real-time inputs. Developers can use channels to forward notifications such as CI results, monitoring alerts, or chat messages, enabling the AI to take action autonomously.

Telegram and Discord are the first supported platforms, available as plugins that can be installed within Claude Code. Once configured, users can pair their accounts with the AI agent and restrict access through allowlists, ensuring only authorized senders can interact with the system.

The system supports two-way communication. While incoming messages appear in the developer’s terminal, the AI’s responses are delivered directly through the external platform, creating a seamless chat-like experience.

However, the feature requires an active session to function. To enable continuous operation, developers must run Claude Code in a persistent environment, such as a background process.

Toward Always-On AI Agents

The introduction of channels aligns with the growing trend of AI agents acting as continuous collaborators rather than on-demand tools. By integrating messaging platforms, Anthropic is positioning Claude Code as part of a broader ecosystem where AI can monitor, respond, and act across workflows in real time.

The feature also highlights increasing interest in event-driven AI systems. Instead of waiting for user input, these systems can react to external triggers, making them suitable for tasks such as DevOps automation, system monitoring, and collaborative development.

Security controls are a key component of the release. Each channel maintains a sender allowlist, and enterprise users must explicitly enable the feature through administrative settings. This reflects the need to balance automation with controlled access, particularly in team environments.

Anthropic noted that the feature is still evolving, with potential changes to functionality and protocol as feedback is incorporated. For now, channel support is limited to approved plugins, though developers can experiment with custom integrations under restricted conditions.

OpenAI announced plans to acquire Astral, a startup known for its widely used open-source Python developer tools, in a move aimed at expanding the capabilities of its Codex platform.

The acquisition, which remains subject to regulatory approval, will bring Astral’s tooling and engineering team into OpenAI’s ecosystem. The companies said Astral’s products will continue to be supported as open-source projects after the deal closes.

Astral has built a suite of tools that are widely adopted across the Python ecosystem. These include uv for dependency and environment management, Ruff for code linting and formatting, and ty for enforcing type safety. Together, these tools are used by millions of developers to streamline workflows, improve code quality, and reduce errors.

OpenAI said integrating these tools into Codex will allow its AI systems to interact more directly with real-world development environments. The goal is to move beyond code generation and toward AI systems that can participate across the entire software lifecycle.

Expanding AI in Developer Workflows

Codex, OpenAI’s AI system for programming tasks, has seen rapid growth in recent months. The company reported a threefold increase in users and a fivefold rise in usage since the start of the year, with more than two million weekly active users.

The platform is being positioned as a broader development agent rather than a standalone coding assistant. OpenAI aims to enable Codex to plan code changes, modify existing codebases, run development tools, test outputs, and maintain software over time.

Astral’s tools are seen as a key component in that vision because they are already embedded in developer workflows. By integrating them, Codex could gain the ability to operate directly within established development pipelines rather than functioning as a separate interface.

This approach reflects a wider trend in AI development, where systems are increasingly designed to act as collaborators that can execute tasks across multiple stages of production.

Strengthening the Python Ecosystem

Python remains one of the most widely used programming languages, particularly in artificial intelligence, data science, and backend systems. OpenAI said the acquisition will reinforce its commitment to supporting the language and its developer community.

Astral’s tools play a central role in maintaining code quality and project consistency in Python environments. Their integration with AI systems could enable more automated and reliable development processes, especially in large or complex codebases.

The companies indicated that future updates will focus on deeper integration between Codex and Astral’s tooling. This could allow AI agents to manage dependencies, enforce coding standards, and validate outputs as part of a continuous workflow.

For now, OpenAI and Astral will continue operating independently until the transaction is finalized. After closing, Astral’s team is expected to join OpenAI’s Codex group, where it will contribute to building more advanced AI-driven development systems.

The acquisition highlights OpenAI’s broader strategy to embed AI more deeply into professional tools, positioning its models not just as assistants, but as active participants in software creation and maintenance.

The U.S. Department of Defense is facing internal resistance as it moves to phase out artificial intelligence tools developed by Anthropic, following a decision to classify the company as a supply-chain risk.

Defense Secretary Pete Hegseth issued the designation on March 3 after a dispute with Anthropic over usage guardrails for its AI systems. The order bars the Pentagon and its contractors from using Anthropic’s technology, including its widely adopted Claude model, with a six-month transition period.

However, military personnel, IT staff, and contractors say the directive is proving difficult to implement. Many users have grown reliant on Anthropic’s tools and view them as more effective than competing systems. Some are delaying compliance, while others expect the ban may eventually be reversed.

Operational Dependence on AI Tools

Anthropic’s AI systems have become embedded in military workflows, supporting tasks ranging from data analysis to operational planning. Claude was the first AI model approved for use on classified Pentagon networks, and adoption expanded rapidly following a $200 million defense contract awarded in 2025.

Users say the tools significantly improved efficiency, particularly in handling large datasets and automating repetitive processes. In some cases, developers relied on Anthropic’s Claude Code system to generate software and build automated workflows.

With the phase-out underway, some of these processes are reverting to manual methods. One official said tasks previously handled by AI, such as querying large datasets, are now being performed using traditional tools like spreadsheets, resulting in slower workflows and reduced productivity.

Replacing Anthropic’s systems is also technically complex. Contractors note that recertifying alternative AI models for use on classified networks could take between 12 and 18 months. This process includes rigorous security and compliance checks, making a rapid transition unlikely.

Cost, Complexity, and Strategic Uncertainty

The removal of Anthropic’s tools is expected to carry both financial and operational costs. Systems built around Claude may require partial redesign, particularly in cases where workflows and prompts were tailored to its architecture.

For example, software platforms used for intelligence analysis and targeting operations rely on AI-driven workflows that would need to be rebuilt using alternative models. Contractors say this process could delay projects and reduce efficiency in the short term.

At the same time, Pentagon officials and contractors are weighing whether to fully transition to other providers, such as OpenAI, Google, or xAI, or to adopt a more gradual approach. Some agencies are reportedly slowing their phase-out efforts in anticipation of a potential resolution between the government and Anthropic.

The situation highlights a broader challenge in AI adoption within government systems: balancing security concerns with operational effectiveness. As AI tools become more deeply integrated into critical workflows, replacing them can create significant disruption.

The Pentagon’s experience underscores how quickly AI technologies have moved from experimental tools to essential infrastructure. It also illustrates the growing tension between policy decisions and the practical realities of deploying advanced AI systems at scale.