# Anthropic details safety, privacy, and security for Claude Duet agent

> Original title: Claude Duet Safety, Privacy, and Security

- Source: Claude Blog
- Published: 2026-10-23T04:00:00.000Z
- Bad Influence: https://badinfluence.news/items/gg3hhpay1gchgvmkh079ikgps
- Original: https://claude.com/resources/articles/claude-duet-safety-privacy-and-security

## Summary

Anthropic describes safety, privacy, and security protections for Claude Duet, a research preview agent that runs continuously on a cloud computer and uses connected apps. Before sending sensitive messages, paying, or deleting anything, the agent shows the action and waits for user approval. A separate classifier checks outgoing messages, and passwords and sign-in tokens never reach the model.

## Why it matters

The post explains how Claude Duet keeps credentials away from the model and requires approval before sensitive actions, which matters for anyone weighing a continuously running agent.