product updateAnthropic

Anthropic Unifies Claude's Chrome Side Panel with Cowork, Syncing Browser Sessions Across Apps

TL;DR

Anthropic has upgraded Claude's Chrome side panel into a full Cowork session, letting browser-based conversations and tasks carry over to Claude's desktop, web, and mobile apps. The feature is live now for Max and Team customers, with Pro access rolling out in coming weeks.

3 min read
0

Anthropic has turned Claude's Chrome side panel into a full Cowork session, allowing browser-based tasks to persist across the company's other apps rather than remaining siloed in the browser.

What changed

Announced today, the update saves side panel conversations to Claude history, making them resumable from Claude's desktop, web, and mobile apps. Previously, side panel sessions existed separately from the rest of Claude's ecosystem — starting a task in Chrome meant losing that context if you switched to another surface.

Skills and connectors already configured on a user's account now work directly from the browser without additional setup.

How it works

Claude in Chrome can view the current page and act on it using the user's existing browser logins — clicking links, navigating between pages, entering text, and filling out forms. Anthropic's own example describes asking Claude to pull invoice details from multiple vendor portals, compile the data into a spreadsheet, then picking up that same session on desktop to incorporate local files or compare figures against a previous budget.

This differs from Claude Code's in-app browser, which Anthropic highlighted recently and which runs on an isolated profile. The Chrome extension instead operates within sites where the user is already authenticated, raising the stakes for security.

Security safeguards

Anthropic acknowledges that browser agents remain vulnerable to prompt injection attacks. To mitigate this, a separate review step checks consequential actions when automatic approval is turned on, and Claude still requires explicit confirmation before purchases or sharing personal data. The company states these measures reduce — but do not eliminate — risk.

Availability

The Cowork side panel is live today for Max and Team subscribers. A Pro tier rollout is planned over the coming weeks. Enterprise administrators can enable the feature and restrict its use to approved domains.

Claude in Chrome does not yet support other Chromium-based browsers or mobile devices, according to Anthropic.

The update fits a broader pattern at Anthropic of making Claude's work portable across surfaces. Recent related additions include live iOS app testing in Claude Code for Mac and new Opus 4.8 controls aimed at longer-running tasks across more products and devices.

What this means

This is a product integration change, not a new model release — no new weights or model version is involved. The significance lies in workflow continuity: browser automation has historically been a dead end for context, forcing users to restart or manually transfer information when moving from Chrome to desktop or mobile. Syncing session state across surfaces addresses a real friction point for anyone using Claude as an agent to complete multi-step tasks that span web research and local file work.

The prompt injection caveat is worth taking seriously. Any agent that can act inside a logged-in browser session — clicking, typing, navigating — inherits the security exposure of the sites it operates on. Anthropic's confirmation-before-purchase safeguard is a reasonable baseline, but the company's own admission that risk can be reduced rather than eliminated suggests enterprises should treat domain restrictions as a requirement, not an option, before deploying this broadly.

Related Articles

research

Anthropic Report: Claude Was Used to Target US Navy Ships, Build Missiles, and Track Uyghurs

Anthropic's latest threat intelligence report documents five cases where state and non-state actors used Claude for military targeting, weapons development, mass surveillance, and repression. The findings include an Iran-linked operation targeting US naval forces and a Mali-based system capable of monitoring 25 million phones.

analysis

Anthropic Threat Report: Claude Used for Missile Software, Mass Surveillance, and Systematic Theft by Chinese AI Labs

Anthropic's latest threat intelligence report covers December 2025 through August 2026, documenting Claude's misuse in espionage, weapons development, and nationwide surveillance operations. The report also details how seven Chinese AI labs ran covert networks—some routing their own customers' requests through Claude—to extract training data at industrial scale.

analysis

Analysis: Claude 'Fable 5.1' Drops Em Dashes and Hedging Language, Answers Grow 30% Longer

A new Arena.ai analysis of tens of thousands of Text Arena outputs shows Claude 'Fable 5.1' has shifted its writing style significantly from Fable 5 — using fewer em dashes, less hedging language, and producing 30% longer responses. The codenamed models appear to be unreleased Anthropic checkpoints being tested anonymously on LMArena.

analysis

Anthropic CEO Dario Amodei Proposes Three-Step Plan to Deliberately Slow AI Capability Advances

Anthropic CEO Dario Amodei published an essay proposing a three-step plan to deliberately pace AI development, including third-party safety audits and cross-industry coordination. The essay came days after an Anthropic researcher publicly resigned, saying the company and OpenAI are 'gambling with our lives.'

Comments

Loading...

Claude Chrome Side Panel Now Full Cowork Session | TPS