Anthropic appears to be blocking the term "OpenClaw" within Claude Code subscriptions. Users hitting this restriction get told to switch to API access, according to a screenshot making the rounds online. It's automatic, triggering whenever the term shows up in context. Nobody seems to know what "OpenClaw" actually is. Quick searches turn up no GitHub repo, no paper, no widely-known tool by that name. Anthropic hasn't explained the restriction publicly. Since subscription and API access have different pricing and rules, this creates a real friction point for anyone whose work involves this mystery term. Blocking a specific word like this is odd. Content moderation usually targets harmful or illegal material, not random product names. Without official comment, we're left guessing: competitive tool? Technique Anthropic wants distance from? Something else entirely? Whatever the answer, it's a reminder that subscription AI products come with guardrails that may not apply to API access.