Analysis
Anthropic expanded Claude Cowork, its agentic task-management product, from a desktop-only app to mobile and web starting July 8 -- initially for subscribers on the Max plan, with broader plan availability rolling out over the following weeks.
The expansion lets users start a task from their desk, get status updates on their phone, and pick up finished output later even with their laptop closed. Cowork sessions now run remotely in beta, saved to the user's Claude account rather than tied to a single device, and chat and Cowork now share one home tab with a single sidebar and search across web and desktop.
“Anthropic paired the launch with usage data from 1.2 million anonymized Cowork sessions sampled across more than 600,000 organizations over two weeks in May.”
Anthropic paired the launch with usage data from 1.2 million anonymized Cowork sessions sampled across more than 600,000 organizations over two weeks in May. The single largest category, at 33.4% of usage, was general business-process work -- pulling scattered updates into a report, building onboarding checklists, reconciling spreadsheets -- versus just 8.7% for software development, undercutting the narrative that agentic AI tools are primarily a coding phenomenon.
The timing lines up with Anthropic's broader competitive posture this week: the company also extended doubled Cowork usage limits through August 5, a retention move that lands in the same 48 hours it separately reversed course on ending free Claude Fable 5 access, just as OpenAI's GPT-5.6 clears for a global public launch. For enterprise buyers, the data point worth internalizing is that agentic AI adoption inside companies is already broader than the developer-tools framing most vendors lead with.