Anthropic has upgraded its Claude AI Chrome extension, allowing users to save chat threads across devices and streamline workflows within its Cowork system, marking a significant step in AI-enabled cross-platform productivity.
Anthropic has updated its Claude in Chrome extension so that chats started in Chrome’s side panel are now saved to Claude history and can be resumed across the desktop app, web and mobile app. The change removes a long-standing limitation for users who relied on the browser panel for short tasks but lost the thread once the window closed. According to Anthropic’s announcement, the browser experience is now tied into Cowork, the company’s broader cross-device workflow system.
Cowork has been evolving quickly this year. TechCrunch reported in January that Anthropic launched it as a research preview for Max subscribers, positioning it as a way for non-programmers to use Claude to work inside selected folders on their computers. By July, the same publication said Cowork had expanded to mobile and web, letting users begin work on a desktop device and pick it up later elsewhere. The Chrome integration extends that model into the browser, where Claude can see the open page, follow links, move between pages, type text and complete forms using the user’s existing logins.
Anthropic says the extension can also use skills and connectors already set up on an account without extra configuration. That makes it more useful for routine administrative work, such as gathering invoice data from supplier portals, building a spreadsheet and then continuing the same session on a desktop to attach local files or compare against an earlier budget. The company’s support pages say Claude in Chrome is available on paid plans including Pro, Max, Team and Enterprise, while administrators on enterprise accounts can restrict access to approved domains.
The added convenience comes with familiar safety concerns. Anthropic warns that browser agents remain exposed to prompt injection, a technique in which malicious page content tries to steer an AI system into unwanted actions. The company says it has added separate checks for higher-risk actions when automatic approval is enabled and still asks for confirmation before purchases or the sharing of personal data. That reduces risk, but does not remove it entirely.
Disclaimer: This content is intended for informational purposes only. Readers are advised to exercise their own judgement, conduct due diligence, or consult a qualified expert before acting on any information provided.
