OpenAI debuts always-on agents to end the friction of manual team handoffs

OpenAI introduced always-on workspace agents in ChatGPT for teams, designed to run in the cloud and handle multi-step work across connected tools and organizational data.

The agents can take on repeatable tasks such as reporting, request triage, and messaging, pushing ChatGPT deeper into enterprise workflows.

OpenAI added enterprise controls including permissions, approval steps, and admin monitoring to support safer deployment.

Custom GPTs remain available, and OpenAI plans migration paths from custom GPTs to agents over time.