OpenAI announces workspace agents in ChatGPT, letting teams create Codex-powered shared agents for complex tasks, and says they are “an evolution of GPTs”
Codex-powered agents for teams. — Create an agent(opens in a new window)Contact sales(opens in a new window)
Context & Ripple Effects
OpenAI’s Codex story has widened from a software-engineering agent for paid ChatGPT business tiers to a system the company describes as capable of handling a much broader set of computer-based professional work.
This release brings that capability into a shared team construct inside ChatGPT, while later coverage of a unified ChatGPT and Codex desktop app reinforces the direction: fewer separate entry points for chat, coding, and work-oriented agents.
First-order effects
- Teams can create and share Codex-powered agents in ChatGPT for complex work, shifting agent setup from an individual GPT-style configuration toward a workspace asset.
- OpenAI makes ChatGPT a more direct delivery surface for Codex capabilities, rather than keeping advanced agent work primarily associated with a separate coding product.
Second-order effects
- Organizations evaluating Codex can standardize reusable team agents inside the ChatGPT environment, increasing the value of their existing ChatGPT deployment relative to one-off individual assistants.
- The move raises pressure on workplace AI products to combine agent creation, shared use, and task execution in the same collaboration surface rather than treating chat and automation as separate tools.
Third-order effects
- If this rollout becomes the common model, workplace AI competition will increasingly center on the workspace that owns reusable agents and their day-to-day use—not only on the underlying model or a standalone coding agent.
- The pattern points to agent capabilities generalizing from specialist development workflows into configurable team workflows, though the corpus does not establish how broadly teams will adopt or govern those agents.
The trend: This is one step in the consolidation of chat, coding, and shared task agents into a single workplace AI surface.