Anthropic Introduces AI Agent Claude for Chrome

Anthropic has announced the launch of Claude for Chrome, a browser-based AI agent available as a limited research preview. Revealed on August 26, the tool will initially be accessible to only 1,000 subscribers of Claude's Max plan, priced between $100 and $200 per month.

The agent enables users to interact with Claude in a side window that retains browser context, facilitating smoother task delegation while browsing. With user permission, Claude can also perform certain actions within the browser on their behalf, marking progress toward AI systems capable of handling more complex online tasks.

Anthropic stated that it has implemented safeguards against prompt injection attacks and will use this preview to test and address emerging safety risks.

The competition to integrate AI into browsers has intensified recently. AI companies view browser-based assistants as the next major battleground for user adoption. Anthropic's rollout coincides with Google awaiting a final ruling in an antitrust case, where a federal judge suggested Google may need to sell its Chrome browser. This uncertainty has sparked interest from competitors: Perplexity made a surprise $34.5 billion offer for Chrome, while OpenAI CEO Sam Altman has also shown interest in acquiring it.

Meanwhile, Perplexity has already launched its Comet browser with AI task delegation features, and OpenAI is reportedly developing its own AI-powered browser with similar capabilities. Google has begun integrating Gemini into Chrome in recent months.

Anthropic emphasized that users will remain in control of the new browser agent. Access to certain categories, such as financial services, adult content, and pirated material, is blocked by default. For high-risk actions, including publishing, purchasing, or sharing personal information, the AI will request explicit user consent. Users can also limit Claude's access to specific sites directly through the app's settings.