Anthropic Cuts Claude Pro/Max Subscribers Off from OpenClaw and Third-Party Agent Frameworks in Cost Crackdown

Anthropic has stopped allowing Claude Pro and Max flat-rate subscribers to use third-party AI agent frameworks such as OpenClaw, effective April 4, 2026, requiring users to pay under a new 'extra usage' billing system to continue. The move is aimed at alleviating strain on Anthropic's infrastructure and prioritizing usage within its own Claude Code environment. The decision generated 761 comments on Hacker News (994 points), with community debate split between understanding Anthropic's business sustainability concerns and frustration over perceived vendor lock-in tactics. Anthropic separately announced a partnership with the Australian government to strengthen AI safety frameworks and track economic impacts of AI deployment across industries.