The Great AI Harness Crackdown: When LLM Vendors Decided You're Using Their Models Wrong
In early 2026, Anthropic and Google began banning developers who used third-party coding tools like OpenCode and OpenClaw with their APIs. What started as quiet rate limiting escalated into legal threats, mass account bans, and deleted forum posts — all driven by vendor paranoia about model distillation and token arbitrage. A firsthand account of bewilderment.
AI LLM Developer Experience OpenCode Anthropic Google Open Source