OpenAI Launches Workspace Agents: Autonomous AI That Runs in the Cloud
New Codex-powered agents can automate complex workflows across tools, marking OpenAI's push into enterprise automation.
Crédit photo: Photo via Unsplash. Free to use under Unsplash License. · source
OpenAI has introduced workspace agents in ChatGPT, a new capability that lets teams deploy autonomous AI agents to handle complex, multi-step workflows without constant human oversight.
What are workspace agents?
Workspace agents are Codex-powered AI systems that run in the cloud rather than on your local machine. Unlike standard ChatGPT conversations where you ask questions and get answers, these agents can take action. They connect to external tools, execute tasks across multiple platforms, and complete workflows that would normally require a human to click through several applications.
Think of them as digital assistants that can actually do the work, not just tell you how to do it.
How do they differ from regular ChatGPT?
The key distinction is autonomy and persistence. A regular ChatGPT session ends when you close the window. Workspace agents, by contrast, can run independently in the cloud. You set them up, define what they should accomplish, and they continue working even when you step away.
They also operate across tools. Rather than being confined to a chat interface, these agents can connect to the software your team already uses, moving data and triggering actions between systems.
Who is this built for?
À lire aussi
More in AI Models
The new Trusted Access for Cyber initiative provides vetted security professionals with specialized AI models and $10 million in API grants, while attempting to keep the same capabilities away from attackers.
Mark Kowalski · 2 hours ago · 2 min
The company has published detailed tutorials for setting up workspaces, managing projects, and completing tasks with its software engineering assistant.
Aisha Patel · 2 hours ago · 2 min
The partnership will let companies run Codex, OpenAI's software engineering agent, on their own infrastructure rather than in the cloud.
James Chen · 2 hours ago · 3 min
A breakthrough in sim-to-real transfer could dramatically accelerate how quickly robots learn new tasks, eliminating the need for costly real-world training data.
