Anthropic Temporarily Bans OpenClaw Creator from Claude Over Pricing and Suspicious Activity

Recently, AI model company Anthropic temporarily banned Peter Steinberger, the creator of the open-source tool OpenClaw, from accessing its Claude platform due to "suspicious" activity. Steinberger announced the account suspension on social media and expressed concerns that future compatibility between OpenClaw and Anthropic's models would become more difficult. This ban followed Anthropic's policy adjustments, sparking industry attention on API control and developer relations.

Last week, Anthropic stated that its Claude subscription would no longer cover "third-party tools including OpenClaw," a move developers have dubbed the "claw tax." The company launched its own intelligent agent, Cowork, adding remote control features similar to OpenClaw. Steinberger claimed he had complied with the new rules using the API but was still banned, questioning whether Anthropic copied open-source features into a closed system before restricting external developers. Though not explicitly stated, this may refer to features like Claude Dispatch added to Cowork.

Although Anthropic later restored Steinberger's access, this incident highlighted friction between AI platforms and third-party developers as API policies tighten. As AI companies increase control over technology use, balancing open-source ecosystems with commercial interests will be a key challenge, potentially shaping future innovation and collaboration.