Goldman Sachs Restricts Claude Code for Bankers and Traders
Goldman Sachs has restricted access to Anthropic's Claude Code for its bankers and traders due to regulatory, compliance, and cost considerations. However, the firm continues to maintain its own internal multi-model AI assistant across the organization. This move highlights the growing enterprise friction and strict risk management surrounding third-party autonomous AI coding agents in highly regulated sectors like finance. It illustrates why major financial institutions are hesitant to adopt consumption-based billing tools that operate with minimal oversight. The restriction specifically targets third-party agents capable of writing and executing code independently, particularly affecting traders who possess permissions to send market orders and book client transactions. Meanwhile, Goldman's custom multi-model internal assistant remains available for general use.
## BACKGROUND
Claude Code is an agentic coding tool developed by Anthropic that operates directly within terminal and IDE environments to edit files, run commands, and manage codebases. Financial institutions face unique compliance challenges because autonomous coding agents acting under a user's account can potentially trigger operational or regulatory risks if left unmonitored.