Introducing workspace agents in ChatGPT
New cloud-based agents use Codex to scale team tasks securely across multiple tools.
OpenAI has introduced a significant new capability for its ChatGPT platform called Workspace Agents. Powered by the company's advanced Codex model, these agents are designed to move beyond simple conversation and execute complex, multi-step workflows. They operate in the cloud, which allows for more powerful processing and persistent task management, and are built to integrate securely with a team's existing suite of tools, from data platforms to communication apps.
This development marks a shift from ChatGPT as a reactive assistant to a proactive, automated workforce. Teams can now configure agents to handle intricate processes like compiling weekly performance reports from multiple data sources, managing project timelines across different software, or orchestrating customer onboarding sequences. The "secure" operation is a key selling point for enterprise adoption, addressing data privacy and governance concerns that have previously limited AI tool integration in professional environments.
- Powered by OpenAI's Codex model for advanced reasoning and code execution.
- Automates complex, multi-step workflows instead of single-turn tasks.
- Runs in the cloud and integrates securely across a team's existing toolset.
Why It Matters
Enables teams to automate entire business processes, moving from AI assistants to autonomous digital workers that scale operations.