Anthropic Briefly Bans OpenClaw Creator Peter Steinberger — Then Reinstates Him Hours Later

Peter Steinberger, creator of OpenClaw, was briefly suspended from Anthropic’s platform Friday morning — the latest twist in an already contentious relationship between the viral open-source AI agent framework and the AI company whose models it runs on.

What happened: Steinberger posted a screenshot showing his Anthropic account had been suspended over “suspicious activity.” Within hours, after his post gained traction, an Anthropic engineer publicly reached out offering help — and his account was reinstated. Anthropic has not publicly commented on what triggered the ban.

Context: Last week, Anthropic announced that subscriptions to Claude would no longer cover “third-party harnesses including OpenClaw,” effectively instituting what developers have dubbed a “claw tax.” Users must now pay for that usage separately via Claude’s API, based on consumption. Steinberger said he was following the new rules and using his API key — but was banned anyway.

The beef with Anthropic: Steinberger is now employed by OpenAI, a direct Anthropic competitor. He has been publicly critical of the company’s decisions, posting: “Funny how timings match up, first they copy some popular features into their closed harness, then they lock out open source.” He declined to elaborate on specifics, but observers noted Anthropic’s recent rollout of “Claude Dispatch” — which lets users remotely control agents and assign tasks — preceded the OpenClaw pricing change by about two weeks.

When asked why he uses Claude at all given his employer, Steinberger said it’s strictly for testing: “My work [at OpenAI] doesn’t mean I should leave Claude users of OpenClaw hanging.”

The episode underscores the growing friction between open-source agent frameworks and the model providers whose ecosystems they depend on. OpenClaw now sits at 347,000 GitHub stars and has been the subject of two major security advisories in the past month alone.