OpenAI on Thursday unveiled ChatGPT Work, an agent in its popular chatbot designed to perform tasks across various applications and files, marking the startup’s latest push into workplace automation.
Powered by OpenAI’s advanced AI model GPT-5.6, ChatGPT Work can gather context from apps, files and workflows to create finished documents, spreadsheets, presentations, reports and websites, the company said.
The launch reflects heightened competition to build and sell AI tools for professional use, as tech companies seek to capitalize on growing demand for autonomous agents that can perform complex tasks with minimal human input.
Prior to this launch, OpenAI’s agent offerings included operator and deep research, later consolidated into the ChatGPT Agent for individual users, as well as Workspace Agents for automating enterprise workflow.
ChatGPT Work comes months after OpenAI competitor Anthropic stepped up its enterprise efforts with Claude Cowork, an agent capable of autonomously planning and executing multi-step tasks.
Earlier this year, Anthropic launched plug-ins for its Claude Cowork agent to automate legal, sales, marketing and data analytics tasks. The move triggered a sharp sell-off in US and European software and professional services stocks on fears of widespread AI-driven disruption of the data analytics industry.
Shortly after Anthropic, Microsoft MSFT.O, an OpenAI backer, also unveiled Copilot Cowork to build out its agentic AI offering.
OpenAI, which is preparing for an IPO, also announced a new ChatGPT desktop application and a hosted website feature that allows users to build and share websites directly through ChatGPT Work.
ChatGPT Work is rolling out Thursday on web and mobile, starting with Pro, Enterprise and Edu users, and expanding to Plus and Business users over the next few days.
