Anthropic has announced a significant update for its popular AI chatbot, Claude. The company aims to change the user experience by adding a dedicated web browser directly integrated into the Claude desktop application.

This new feature allows Claude to automatically detect commands requiring an internet search and display the results in a side panel within the application. Called Claude Cowork, this system aims to reduce the need for users to switch between browsers.

How Does Claude’s New Browser Feature Work?

According to information shared by Anthropic, the new browser is designed only to search publicly available information. Users’ email, banking, or other online service login information is not accessible through this browser.

This restriction is highlighted as part of a security-focused approach. Instead of manually using the browser, users can allow Claude to automatically activate when needed.

This feature is currently active for users with Enterprise plans. Pro, Max, and Team subscribers will also be able to access the new browser this week.

This feature, which is enabled by default, allows users or system administrators to continue using Claude through Chrome via the settings. With this step, the company plans to transform Claude from a chatbot that only answers questions into an assistant that can manage digital missions.

Google continues to manage approximately 70 percent of internet traffic through Chrome. Anthropic’s move is seen as a strategy to break Google’s dominance by directly integrating users’ internet browsing habits into chatbots.

Security Risks and Future Expectations

Direct access of artificial intelligence models to the internet also brings with it a number of technical risks. Anthropic acknowledges that prompt injection attacks, which are hidden in web pages and aim to direct AI spies, pose a risk to its new browser. The company states that the same security measures used in the Claude in Chrome extension are also valid against these types of attacks. However, experts in the field believe that AI spies’ vulnerability to these types of attacks is a difficult problem to solve. OpenAI followed a similar strategy, previously shutting down its independent Atlas browser and adding a built-in browser to ChatGPT. These developments clearly demonstrate the desire of AI companies to manage users’ entire digital lives from a single window. Whether AI spies with internet access can operate reliably remains a subject of debate in the technology world. How do you think AI chatbots having their own browsers will affect our internet usage habits?

Anthropic enables users to perform internet searches from within the application with its new web browser integrated into the Claude chatbot.

Browser, Claude, User, Artificial Intelligence, Internet

<

div>