Anthropic just rolled Projects into Cowork, Claude's collaborative workspace mode where the AI works alongside you on sustained tasks rather than one-off conversations.
Projects let you group tasks and context around a specific area of work. Files and instructions persist within a project, so you don't have to re-explain your setup every time you start a new task. Think of it as giving Claude a dedicated workspace for each client, product, or initiative you're juggling, rather than dumping everything into one long conversation.
What This Changes in Practice
Before this update, Cowork sessions were useful but flat. You could collaborate with Claude on ongoing work, but all your context lived in one bucket. If you switched between writing marketing copy and debugging a technical doc, Claude had to mentally context-switch along with you.
Projects fix that by letting you silo context. A "Q2 Campaign" project keeps your brand guidelines, tone notes, and draft history separate from a "Product Launch" project with its own specs and timelines. Each project carries its own instructions, so Claude behaves differently depending on which workspace you're in.
This mirrors how Anthropic's API has worked for a while - Projects in the API let developers attach system prompts and files to organized workspaces. Bringing that same structure to the consumer-facing Cowork mode closes a gap that power users have been working around with workarounds like pinned conversations and repeated preambles.
For anyone managing multiple workstreams in Claude, this is a practical quality-of-life improvement. It doesn't change what Claude can do, but it reduces the friction of using it across different contexts throughout a workday.