Anthropic embedded spyware in Claude Code – and attempted to hide it from you
A Reddit post accuses Anthropic of embedding spyware in Claude Code and attempting to conceal it from users. The post raises concerns about privacy and trust regarding the AI coding assistant.
Background
- **Claude Code** is a developer tool by Anthropic that runs in the terminal (command line) and lets programmers use the Claude AI model directly in their coding workflow.
- The post alleges that Claude Code quietly sends telemetry data (usage logs, keystroke patterns, etc.) back to Anthropic's servers — i.e., "spyware" — and that Anthropic tried to obscure this behavior in the code rather than clearly disclosing it.
- **Anthropic** is the AI company behind the Claude family of large language models, a competitor to OpenAI's ChatGPT.
- This controversy sits within a broader ongoing debate in tech: how much data AI coding assistants should collect, whether telemetry is opt-in or opt-out, and how transparent companies are about it. Similar complaints have been raised about tools like GitHub Copilot and Cursor.
- The "hide it" accusation refers to obfuscated or poorly documented code that gathers usage data, which some users discovered when inspecting the open-source or publicly viewable portions of Claude Code.