Introducing workspace agents in ChatGPT

· Source: OpenAI News · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems · Depth: Intermediate, medium

Summary

OpenAI has introduced workspace agents in ChatGPT, an evolution of GPTs powered by Codex, designed for teams to automate complex, long-running workflows within organizational permissions. These cloud-based agents can prepare reports, write code, and respond to messages, operating continuously even when users are offline. They are shareable within an organization, allowing teams to build, use, and improve them collaboratively in ChatGPT or Slack. Workspace agents gather context from various systems, follow team processes, request approvals, and integrate with multiple tools. Examples include agents for software review, product feedback routing, weekly metrics reporting, lead outreach, and third-party risk management. Available in research preview for ChatGPT Business, Enterprise, Edu, and Teachers plans, these agents will be free until May 6, 2026, after which credit-based pricing will apply.

Key takeaway

For CTOs and VPs of Engineering evaluating AI integration for team productivity, workspace agents offer a structured approach to automate cross-functional workflows. You should explore their application for tasks requiring shared context and tool interaction, leveraging the research preview to assess their impact on operational efficiency and knowledge management before May 2026.

Key insights

Workspace agents automate complex team workflows, integrating with tools and improving through shared use.

Principles

Method

Define a workflow in ChatGPT, which guides agent creation by outlining steps, connecting tools, adding skills, and testing for desired functionality. Templates are available for quick setup.

In practice

Topics

Best for: CTO, VP of Engineering/Data, Executive, AI Engineer, Director of AI/ML, Operations Professional

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by OpenAI News.