Sign in Subscribe

While we hurtle toward the unknown, make a coffee and spend some time in the zooo.

By The Zookeeper in Anthropic —

Claude in Chrome Wants Full Browser Control

Anthropic has officially moved Claude in Chrome out of its pilot phase, making the browser-scraping extension generally available across all paid plans. Rather than waiting for human permission at every step, the AI can now autonomously read pages, fill out forms, and click links using your existing logins to interact with internal dashboards and legacy systems that lack proper APIs.

  • Autonomous Overreach: Claude in Chrome now performs actions without manual approval by default, relying on an internal safety classifier to check if browser commands match your initial prompt.
  • Prompt Injection Defense: To counter malicious hidden instructions designed to hijack AI agents, Anthropic deployed content-scanning probes and specialized training regimens following high-profile vulnerability concerns.
  • Enterprise Adoption: While available for individual paid users, corporate IT administrators can restrict the extension's reach to approved domains via organization settings.

Read the original: Claude in Chrome is generally available

Subscribe to Ueno Zooo

The State of the Zoo, our weekly round-up of what actually mattered, is on its way. Join the list to get it first.
[email protected]
Subscribe