Anthropic brings Claude Cowork to mobile as 90% of users skip coding

Anthropic has expanded Claude Cowork, its agentic AI tool for general knowledge work, from desktop to web and mobile, rolling out in beta to Max subscribers starting Tuesday. The move frees users from keeping their laptops running to complete tasks, with sessions now executing in the cloud by default.

"Claude Cowork handles the work around the work — the spreadsheets, the reports, the inbox triage — so users can focus on decisions," an Anthropic spokesperson said. "The expansion to mobile means you can start a task at your desk and check results from anywhere."

The company also released early usage data that undercuts the coding-first stereotype of AI agents. Across 1.2 million anonymized sessions from more than 600,000 organizations in late May, business process work such as reconciling spreadsheets and building reports accounted for 33.4% of usage. Content creation and copywriting followed at 16.4%, while software development made up just 8.7%. Anthropic said more than 90% of Cowork sessions involve everyday office tasks rather than programming.

The expansion positions Anthropic against a wave of competitors racing to embed AI agents into enterprise workflows. OpenAI is broadening Codex from a coding tool into a general enterprise-work platform for non-developers, while Google recently launched Gemini Spark as an agentic assistant integrated with Gmail and other office tools. Startup Viktor raised $75 million in a Series A led by Accel to embed an AI coworker inside Slack and Teams. Anthropic has been widening Claude's footprint across office software for months, from integrating with Microsoft Word to a deployment putting Claude in front of 276,000 staff at KPMG.

Under a feature called Dispatch, a single persistent thread routes each request to the right engine — development work to Claude Code, knowledge work to Cowork — and messages back the result rather than every intermediate step. Scheduled tasks now run in the cloud even when devices are offline, and users receive mobile alerts when Claude needs approval or has output ready. The desktop app retains full local file access for deeper work. Anthropic is doubling temporary Cowork usage limits through Aug. 5 to encourage early adoption.

The shift to mobile and cloud execution carries operational risks. Giving a phone remote control of a desktop agent means a manipulated instruction or phishing link could trigger actions on a connected machine. Anthropic's guidance urges users to connect these agents only if they trust every application in the chain, a caution that grows more salient as autonomous helpers proliferate across office environments.

For investors, the usage data signals that the addressable market for AI agents extends well beyond software developers into the broader office workforce. Anthropic, backed by Amazon and Google, has not disclosed Cowork-specific revenue, but the expansion suggests the company sees agentic AI as a primary growth vector. The competitive landscape is intensifying: OpenAI's enterprise push and Google's Gemini integration mean the battle for workplace AI will be won on distribution and reliability, not just model benchmarks. Anthropic's bet is that an agent that handles the administrative grind — and works from any device — can capture the nontechnical users that represent the largest pool of potential subscribers.

This article is for informational purposes only and does not constitute investment advice.