Skip to main content

Claude Code Gets a Built-in Browser, AI Can Now Surf the Web on Its Own

Anthropic just dropped a major update for Claude Code, its developer tool. The headline feature? A built-in browser that lets the AI roam the web on its own.

Developers can now ask Claude to fetch technical docs, check issue trackers, or pull up any online resource without leaving the app. No more alt-tabbing between windows. The browser uses a tab-based layout, and you can summon it with a quick keyboard shortcut.

Image

Security First

Giving an AI a browser sounds risky, but Anthropic has layered in safeguards. A classifier screens every write operation on external sites, so Claude can't buy things, create accounts, or solve CAPTCHAs without your explicit go-ahead. The browser runs in a clean, isolated environment and doesn't store login credentials.

Enterprise Controls

For teams worried about compliance, administrators can set a whitelist of allowed websites. If that's not enough, they can disable the browser feature entirely. For users who want Claude to work within their logged-in sessions, Anthropic recommends pairing it with a dedicated Chrome extension.

This update shows how AI agents are getting more capable—and more secure. Claude Code's browser isn't just a gimmick; it's a practical tool that could save developers hours of context-switching.

Key Points

  • Claude Code now includes an integrated browser for independent web browsing.
  • The AI can read, click, and type on external pages, but strict security controls prevent unauthorized actions.
  • Enterprise users can whitelist sites or disable the feature for compliance.
  • A Chrome extension is available for users who want Claude to act within their logged-in sessions.