📫 Subscribe | 🗣️ Hire us to speak | 🤝 Sponsor this newsletter | 🎓 Events

Welcome to the Friday edition of our newsletter. We spend Fridays going deeper into tools and trends related to generative AI (and Tuesdays sharing news updates).

This week, Anthropic shared the results of a study of how Claude Cowork was being used, with their core finding being that most people are using Cowork to complete, as they put it, “the work around the work,” which they describe as consisting of “tasks that are part of a broad swath of jobs, but are rarely a person’s core responsibility.”

In case you missed an earlier Friday discussion of Claude Cowork in our newsletter, Cowork is “Claude Code for ordinary people,” an agentic tool that has access to our files and can do work on our computer through a chat interface instead of through the terminal, which is Claude Code’s point of entry into a user’s computer.

Cowork has been available since January of this year. To understand how it was being used, Anthropic studied Cowork usage during May 2026 across a number of sample sessions.

Read on below to learn more about what they found.

AI is moving faster than any other technology. New models. New tools. New claims. New noise.

Most people feel like they're behind. But the people that don't, aren't smarter. They're just better informed.

The Future Today is a daily briefing for people who want clarity. In one concise email each day, you'll get the most important AI and tech developments, learn why they matter, and what they signal about what's coming next.

Written for operators, builders, leaders, and anyone who wants to sound sharp when AI comes up in the meeting.

One email. Five Minutes. Stay ahead of 99% of the world.

Join Free Today →

To conduct their study, Anthropic sampled Cowork usage across 1.2 million anonymized sessions from May 11-31, 2026, classifying the use cases they found into one of twenty different categories. Significantly, they found that nearly 50% of Cowork usage pertained to two categories, (1) business process and operations and (2) content creation and copywriting.

Tasks falling into the category of business process and operations accounted for 33.4% of Cowork usage, including fairly mundane tasks such as “pulling scattered updates into a single report, building onboarding checklists, and reconciling spreadsheets.” Tasks in this category appear to be the kind of boilerplate, routine, format-intensive tasks that we all encounter in our jobs that don’t require the highest of our intelligence.

The category of content creation and copywriting needs less explanation. Anthropic summarizes it succinctly as tasks that involve “producing drafts, slide decks, posts, and proposals,” a category that accounts for 16.4% of Cowork usage.

Other categories include software development (8.7%), DevOps and infrastructure (7.0%), research and intelligence (6.4%), data analysis and business intelligence (5.8%), document processing and extraction (4.1%), and sales and revenue operations (4.0%). The final twelve smaller categories account for the remaining 14.1% of Cowork usage and include categories such as personal assistance, education, and meeting intelligence.

Here’s a visual breakdown of the top categories by usage:

The main upshot of the study confirms a message that Snider and I regularly deliver at our presentations and workshops, namely that generative AI tools excel at those routine tasks that occupy much of our time but do not constitute “the best parts of our jobs,” i.e., those aspects of our jobs that depend on human expertise, experience, and relationships.

Anthropic captures this point fairly well: “Our data suggests that people are using Claude Cowork to assemble and structure the information they can use to act on their expertise. A lawyer, for example, might use Claude Cowork to handle document formatting and filing, giving them more time to apply their legal judgment to challenging cases. A hiring manager might use Coworker to schedule meetings and synthesize interview feedback, allowing them to spend more time on candidate conversations and evaluating work samples. And a team lead might use Claude Cowork to produce the slide deck that explains a difficult decision, freeing them up to actually make those tough calls.”

The lesson is not that AI replaces expertise, but that it can clear away the administrative and production work that often prevents us from using that expertise well. The real opportunity is to identify the recurring tasks surrounding our most valuable work and hand more of them over to AI.

If you haven’t tried out Claude Cowork, Anthropic has a helpful webpage that covers the basics of getting started with Claude Cowork. Take a look!