No evidence Anthropic changed Claude billing for OpenClaw on April 4

In short: Despite claims spreading online, there is no evidence in the provided reporting that Anthropic made an April 4 policy change that blocks OpenClaw from using Claude subscription limits or forces separate pay as you go billing.

What happened

A report said Anthropic would change its rules on April 4, stopping people from using their Claude subscription limits with “third party harnesses” like OpenClaw, and moving them to separate pay as you go billing.

But the source material provided here says there is no supporting evidence for that claim. It also says there are no mentions of policy updates, user emails, or billing shifts in the latest reports referenced from March 2026 onward.

Instead, the more concrete development is product work. Anthropic has been improving Claude in ways that overlap with OpenClaw, which is an open source tool (a free to reuse toolkit) for building personal AI agents that can be used inside chat apps like WhatsApp, Telegram, and Discord.

Examples include “Claude Code Channels,” which lets people message Claude powered coding bots through Discord and Telegram, not just through desktop or terminal tools. Anthropic has also highlighted “computer use” features, where Claude can control a mouse and keyboard, like a remote helper, with added safety steps meant to reduce data leaks.

Why it matters

Many people use tools like OpenClaw as a simple wrapper around an AI model, like putting the same engine into a different car. If billing rules really changed, it could raise costs overnight for hobbyists and companies. Based on the information provided here, the clearer story is competition and product catch up, not a confirmed shutdown of third party access.

Source: The Verge AI