OpenCode Breach Exposes Users’ OpenAI, Anthropic, and Google API Keys

18 Oct 2026, 12:33 am · 2h ago · 1 min read · OfficeChaiNew

Open-source AI coding tool OpenCode has disclosed a major security breach, warning that unauthorized actors may have accessed third-party API keys provided by its users. The developer platform allows engineers to plug in their custom API keys from leading AI foundation providers including OpenAI, Anthropic, and Google to generate and edit code. Following the unauthorized system intrusion, users are strongly advised to immediately rotate and revoke their existing API credentials across all connected model providers to prevent potential unauthorized access, financial draining, and illicit model abuse via their accounts.