OpenAI has introduced workspace agents within ChatGPT, a new feature designed to help organisations manage complex, multi-step tasks and workflows that can be shared across teams.
The company said the agents operate in the cloud, meaning they can continue working on tasks even when users are offline.
“Teams can now create shared agents that handle complex tasks and long-running workflows, all while operating within the permissions and controls set by their organisation,” OpenAI said in a statement.
OpenAI said workspace agents are an evolution of its existing GPT-based tools and are powered by Codex. They are built specifically for tasks that require coordination across teams, including gathering information from multiple systems, following established workflows, and requesting approvals at key stages.
The agents can write code, prepare reports, and respond to messages. They can also retain information, use connected applications, and execute tasks on a scheduled basis. OpenAI added that the agents are designed to integrate with a range of tools, including Slack, allowing them to slot into workflows where work is already taking place.
Early users have reported significant time savings. “The hard part of building an agent is not the model. It’s the integrations, memory, the user experience. Workspace agents collapsed that work, so one of our sales consultants built, evaluated, and iterated a sales opportunity agent end-to-end without an engineering team. It researches accounts, summarises Gong calls, and posts deal briefs directly into the team’s Slack room,” said Ankur Bhatt, AI Engineering at Rippling.
Reusable Agents With Organisational Controls
OpenAI said teams can build an agent once and reuse it across multiple projects within an organisation, with the aim of turning team knowledge and best practices into repeatable, automated workflows.
The feature has embedded controls that allow organisations to define precisely how agents operate, including which tools they can access and when human approval is required. For sensitive actions, such as sending emails or editing documents, agents can be configured to pause and seek permission before proceeding.
Administrators can manage permissions, monitor usage, and suspend agents if necessary. OpenAI said enterprise-grade oversight tools are also available, including a Compliance API that gives admins visibility into every agent’s configuration and activity.
Workspace agents are currently available in research preview for ChatGPT Business, Enterprise, Edu, and Teachers plans. The feature will remain free until May 6, 2026, after which OpenAI will introduce a credit-based pricing model.
ALSO READ: The Playground is Closed: 10 Hard Truths from the Cisco AI Summit
