Anthropic takes a step further into the era of intelligent assistants by launching Claude, an innovative chatbot, now available as an extension for Google Chrome. Reserved for paying users, this extension promises to transform the way we interact with the web. Discover how this technology integrates into your browser and the security measures in place to protect your data.
The 3 key points not to miss
- Claude is accessible to paying users via a Chrome extension, reducing the previous access cost.
- The extension allows Claude to browse, fill out forms, and automate tasks across different tabs.
- Security measures are integrated to prevent prompt injection attacks and protect user data.
Features of Claude for Chrome
Anthropic has unveiled Claude, its chatbot, as an extension for the Chrome browser, marking an advancement for paying subscribers. This extension allows Claude to interact with websites, fill out forms, and perform scheduled tasks across multiple tabs. By using this technology, users benefit from an intuitive interface that simplifies many common actions on the web.
The cost of the Pro subscription, set at $20 per month, represents a reduction compared to the initial testing phase, where access was limited to more expensive subscriptions. The Claude for Chrome extension is also compatible with other browsers based on Chromium technology, such as Edge, Brave, and Opera, although Firefox and Safari are not compatible at this stage.
Agentic approach and automation
With this extension, Anthropic adopts an AI-focused agentic approach. Unlike systems that merely answer questions, Claude can perform actions that would normally be carried out by the user. This includes navigating through multiple tabs and automating interactions with websites, thanks to its Opus 4.5 model.
This ability to automate tasks can significantly improve efficiency for users, allowing them to focus on more complex or creative activities. However, this level of integration raises security concerns that the company takes very seriously.
Security and user control
Anthropic is aware of the potential risks associated with the deep integration of Claude into the browser. Threats, such as prompt injection attacks where malicious third parties might attempt to influence the chatbot’s actions, are taken into account.
To minimize these risks, users have the ability to control Claude’s access to each website. Settings can be adjusted at any time, and the extension requires explicit permission for actions deemed sensitive, such as posting content or sharing personal data. Furthermore, Claude automatically deactivates on sites deemed risky, such as banking platforms or those offering illegal content.
Competition and market alternatives
Anthropic is not the only player exploring this field. Earlier this year, OpenAI launched ChatGPT Atlas, a similar AI feature for the browser, also aimed at subscribed users. For those who prefer not to subscribe, Perplexity Comet offers a free alternative since October.
These developments show a growing trend towards the integration of intelligent assistants into browsers, providing users with powerful tools to manage and automate their online interactions.
Anthropic: a company at the forefront of AI
Founded in 2021 by former members of the OpenAI team, Anthropic has quickly positioned itself as a major player in artificial intelligence. The company focuses on developing ethical and safe AI models, with particular attention to the societal implications of these technologies. Through initiatives like Claude, Anthropic continues to push the boundaries of AI while placing security and user experience at the heart of its priorities.