30 Days Gen AI Risk Trial -Start Now
Book a demo
Coding·Bundled with Claude Pro $20/mo, Claude Team $30/seat/mo, or Enterprise (custom)·claude.com

Claude Code

Anthropic's official CLI for Claude — terminal-based agentic coding assistant.

Risk Score
Low
3/10

Independent assessment across data handling, compliance, security and transparency.

Overview

Claude Code is Anthropic's command-line interface for Claude, designed for agentic software-engineering workflows. Runs locally in the user's terminal and connects to the Anthropic API. Conversation data is governed by Anthropic's commercial terms when authenticated through a paid plan; the API does not use prompts/completions to train Anthropic's models by default.

Risk factors

2
  • Source code from the developer's working directory may be transmitted to Anthropic for inference
  • Locally-stored conversation logs and tool-call history accumulate on the developer machine

Recommendations

4
  • Use a paid Anthropic plan (Pro / Team / Enterprise) so commercial terms apply — not the free consumer terms
  • Configure project-level allow/deny lists for tools and files via .claude/settings.json
  • Avoid running Claude Code with --dangerously-skip-permissions on shared / production machines
  • For regulated workloads, use the Enterprise plan with the signed DPA + zero-retention add-on

Data handling

Storage
Inference on Anthropic infrastructure (AWS); conversation logs stored locally on the developer machine
Retention
30 days for API logs by default; zero-retention available on Enterprise
Training on inputs
Off by default — Anthropic does not train on commercial API or product data