Anthropic Launches Dispatch While Cutting Off OpenClaw Access
2 min readAnthropic Ships Dispatch — and Shuts the Door on OpenClaw
In a move that’s equal parts competitive strategy and capacity management, Anthropic this week launched Dispatch — its own take on the remote AI agent experience that tools like OpenClaw have built a following around — while simultaneously cutting off OpenClaw and similar third-party tools from standard Claude subscription access, effective April 4.
What Is Dispatch?
Dispatch is currently a research preview within Anthropic’s Claude Cowork platform. It lets users remotely control a Mac desktop session through their mobile device, with Claude handling the on-screen work. Think of it as giving Claude hands and a screen — accessible from anywhere. Alongside Dispatch, Anthropic also quietly shipped Claude Code Channels, which hooks Claude Code directly into Discord and Telegram so developers can issue coding instructions on the go and get results pushed back to their chat apps.
Both products draw a clear line in the sand: Anthropic wants to own the agentic interface layer, not just power it behind the scenes for third-party tools.
The OpenClaw Cut-Off
The other half of the story is less festive for OpenClaw users. Starting April 4 at 3 PM ET, Claude’s paid subscription plans — including Claude Max — will no longer cover usage routed through third-party tools like OpenClaw and OpenCode. Users who rely on these tools will be moved to pay-as-you-go API billing, which carries higher per-token costs than the flat subscription rate they’ve been enjoying.
Anthropic’s rationale is straightforward: third-party agentic tools generate disproportionately high compute consumption compared to direct chat usage, making the economics of covering them under a flat subscription unsustainable. As a goodwill gesture, affected subscribers will receive a one-time credit matching their monthly subscription price to soften the transition, as reported by TechBuzz AI and VentureBeat.
Why It Matters
The timing here is hard to ignore. Anthropic is launching its own competing products on the same day it cuts off access for the community-built alternatives. Whether you read that as a savvy platform play or an aggressive squeeze on the open ecosystem probably depends on how you feel about Anthropic’s broader trajectory.
What’s clear is that the frontier AI labs are no longer content to be pure model providers. They want the interface relationship with users, and they’re willing to make uncomfortable policy calls to build it. For developers who’ve built workflows on top of OpenClaw, the message is clear: migrate to Anthropic’s own toolchain, or pay up.