Linear Treats AI Agents as Coworkers

This turns Linear from a place where humans track work into a system where software can take work, do work, and report work back. Once agents are treated like named coworkers inside the backlog, Linear can sit in the execution loop for coding, support triage, and product planning, not just the planning layer. That creates new room to monetize agent seats, agent integrations, and higher value automation on top of the core seat business.

• Linear made agents real workspace members in May 2025. They can be assigned issues, added to projects, mentioned in comments, and tracked through their own profiles. That matters because agent work becomes visible inside the same queue, status system, and audit trail teams already use for people.
  1 sacra 2 linear
• The practical use case is not just AI help inside a ticket. It is delegated execution. Linear already supports agents from Devin, Cursor, Codex, Copilot, Warp, and others, while keeping a human as the primary assignee. That makes Linear the control plane where a manager hands work to an agent but keeps accountability with an employee.
  2 linear 3 linear 4 linear
• This model also extends beyond engineering. Linear now uses its own agent in Intercom, Zendesk, and Gong to turn support chats and sales calls into structured issues, then route and de duplicate them in triage. That pushes Linear into support and customer feedback workflows that previously lived outside issue tracking.
  5 linear 6 linear 7 sacra

The next step is straightforward. As coding agents get better at handling multi step tasks, more work will enter Linear already shaped for an agent to execute, and more buyers will judge project management tools by how well they coordinate humans and agents together. That favors products like Linear that already treat agents as operating users inside the system, not just background features.