Claude Code now includes a built-in browser allowing the AI to read, click, and type on external websites directly within the development environment. Write actions are screened by classifiers, with purchases and account creations requiring user approval.
Anthropic has integrated a web browser into Claude Code, expanding the AI's capabilities beyond traditional coding tasks. The feature enables Claude to navigate web pages, extract information, and perform interactions autonomously within the IDE.
The implementation includes safety measures designed to prevent unauthorized actions. A classifier system screens write operations on external sites, flagging potentially risky activities. More sensitive actions—such as making purchases or creating accounts—trigger manual approval workflows, ensuring users maintain control over consequential decisions.
This development represents a significant step toward more autonomous AI assistants. Previously, Claude could only work with code and documents within its environment. Web browsing capability opens possibilities for tasks like research automation, data collection, and testing web applications directly from the coding interface.
The safety-first approach reflects ongoing industry focus on AI alignment. Rather than granting unrestricted web access, the system balances functionality with guardrails. Users can supervise and approve high-impact actions while allowing routine browsing and reading operations to proceed automatically.
For developers, the feature streamlines workflows requiring external data or web interaction. Testing web scraping scripts, verifying API responses, or gathering information for projects can now happen without switching between applications.
The browser integration follows broader trends in AI development—expanding AI capabilities while implementing proportional safety controls. As AI systems take on more autonomous tasks, the pattern of automated routine actions paired with human approval for sensitive operations has become standard practice.
Claude Code users now have access to this feature as part of the platform's ongoing updates.
Uber's weekly AI agent requests have grown nearly tenfold since February, yet the company has held spending flat since April after exhausting its entire 2026 AI budget in Q1.
A recent paper shows artificial intelligence often diagnoses and treats patients better than human physicians. The findings are prompting difficult conversations within the medical community about the profession's evolving role.
The Relay Q, launching next year, represents the latest push to establish voice as the primary interface for human-computer interaction, challenging the keyboard's decades-long dominance.
An Anthropic researcher demonstrated automated systems that can identify and correct misaligned behaviors without compromising overall performance. The systems improved on all 10 tested benchmarks measuring specific problematic outputs.